(Comparative Politics (Oxford University Press) ) Katz, Richard S. - Mair, Peter - Democracy and The Cartelization of Political Parties (2018, Oxford University Press)

OUP CORRECTED PROOF – FINAL, 28/5/2018, SPi
DEMOCRACY AND THE CARTELIZATION

OF POLITICAL PARTIES
COMPARATIVE POLITICS
Comparative Politics is a series for researchers, teachers, and students of
political science that deals with contemporary government and politics. Global in scope,
books in the series are characterized by a stress on comparative analysis and strong
methodological rigour. The series is published in association with the European
Consortium for Political Research. For more information visit
<http://www.ecprnet.eu>.
The series is edited by Emilie van Haute, Professor of Political Science,
Université libre de Bruxelles; and Ferdinand Müller-Rommel, Director of the Center for
the Study of Democracy, Leuphana University; and Susan Scarrow, John and
Rebecca Moores Professor of Political Science at the University of Houston.
OTHER TITLES IN THIS SERIES
Multi-Level Electoral Politics

Beyond the Second-Order Election Model
Sona N. Golder, Ignacio Lago, André Blais, Elisabeth Gidengil, and
Thomas Gschwend
Organizing Political Parties
Representation, Participation, and Power
Edited by Susan E. Scarrow, Paul D. Webb, and Thomas Poguntke
Reforming Democracy
Institutional Engineering in Western Europe
Camille Bedock
Party Reform
The Causes, Challenges, and Consequences of Organizational Change
Anika Gauja
How Europeans View and Evaluate Democracy
Edited by Mónica Ferrín and Hanspeter Kriesi
Faces on the Ballot
The Personalization of Electoral Systems in Europe
Alan Renwick and Jean-Benoit Pilet
The Politics of Party Leadership
A Cross-National Perspective
Edited by William P. Cross and Jean-Benoit Pilet
Beyond Party Members
Changing Approaches to Partisan Mobilization
Susan E. Scarrow
Institutional Design and Party Government in Post-Communist Europe
Csaba Nikolenyi
Representing the People
A Survey among Members of Statewide and Sub-state Parliaments
Edited by Kris Deschouwer and Sam Depauw
Democracy and
the Cartelization of
Political Parties
RICHARD S. KATZ AND PETER MAIR{
1
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Richard S. Katz and Peter Mair 2018
The moral rights of the authors have been asserted
First Edition published in 2018
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2018932389
ISBN 978–0–19–958601–1
Printed in Great Britain by
Clays Ltd, St Ives plc
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
Preface
This book represents the culmination of a project that began at an ECPR

Research Sessions meeting at the University of Mannheim in 1987. For a long
time, it threatened to be one of those “much anticipated forthcoming books”
that somehow never come forth. The project began simply as a data-collection
effort. It appeared obvious to us that there were important changes happening
in the organization of political parties that could have profound consequences
for the way in which democracies work, and that these changes needed to be
understood. But it also appeared obvious that before these changes could
be explained—and at the beginning we had no favored “candidate” to be the
explanation—they simply needed to be recorded in a systematically comparable
way. With the financial support of the American National Science Foundation
(grant SES-8818439) and the Forschungsstelle für Gesellschaftliche Entwicklungen
(FGE) of the University of Mannheim, and with the collaboration of a
talented team of scholars, that is what we set out to do, and in 1992 we
published Party Organizations: A Data Handbook on Party Organizations in
Western Democracies, 1960–90, close to 1,000 pages, almost all of which were
tables, documenting what had happened during those thirty years.
The scope of the project, both geographically and temporally, was largely
determined by practicality. We want to be comprehensive, but limited funds
meant that we could not include every democracy. Moreover, in an era before
elaborate party websites and high-speed internet connections made remote
access of extensive party archives possible, we believed that this research
would require “boots on the ground” in each country included: someone
personally to go to party headquarters and hector party officials until they
delivered the data we wanted. Thus we only included countries for which we
knew, or could readily identify, a local collaborator. The starting date of 1960
was chosen partially in the interest of manageability and partially on the, now
recognized to be dubious, assumption that it would represent a reasonably
stable “old normality” from which change could be assessed. The United
States was included in part to make the project more appealing to the
National Science Foundation (NSF) and in part because it represented a
“different systems” comparator; inclusion of the protoparty system of the
European Union (EU) reflected the strong bias in Europe at the time to
include the EU in any project.
Once the data were in hand and we started our analysis, the question of
scope became more complex. Clearly, we did not intend our conclusions to be
relevant only to the twelve countries (plus the EU party federations) included
vi Preface
in the data-collection effort, but equally we recognized that, as with any
hypotheses that draw on history, or social structure, or institutional arrange-
ments, our conjectures were going to be more appropriate to some times and
places than to others. Even granted that observation, we were never entirely
agreed regarding the appropriate scope for our work. Mair tended to have a
more Euro-centric focus, in particular treating such Europe-specific events as
the Maastricht Treaty as establishing differences between the members of the
EU and those outside of it (and similarly to see the advent of the euro as
establishing differences between countries within the eurozone and those
outside it) that might be seen to limit the scope of our theorizing. Katz, on
the other hand, tended to interpret these events as Euro-specific extreme
examples of more general trends, such that while our conclusions might be
especially relevant within the Eurozone, they were also applicable well beyond
the borders of western Europe.
Indeed, both the idea of a set of cozy arrangements through which osten-
sibly competing parties work together to protect their shared interests, and the
idea that this collusive behavior might be successfully challenged by those
excluded, had roots in the experience of what Katz and Kolodny (1994)
described as a “six-party” national party system of the United States, with
presidential, Senate, and House Democrats, and similarly Republicans, in
many ways organized and acting as three separate, if generally allied, parties.
We saw American politics through the 1970s and 1980s as being characterized
by what we would later call a “cartel” consisting of presidential Demo-
crats and Republicans, Senate Democrats and Republicans, and House
Democrats—but excluding the House Republicans, who had been in the
minority since January 1955, and appeared to be condemned to permanent
minority status. On the one hand, this meant that the other five parties had
little need to accommodate their concerns, and on the other hand it meant
that the House Republicans had little incentive to join with the others in
acting “responsibly.”1 In the end, this led to Newt Gingrich’s “Contract with
America,” which challenged the general elite consensus regarding the charac-
teristics of “responsible” policy, put the Republicans in the majority in the
House of Representatives for the first time in forty years, and by showing that
a frontal attack on “the establishment” and its sense of what was acceptable
could be electorally successful at the national level effectively broke the cartel.
Notwithstanding the significance of this American example, however, our
thinking was largely rooted in the experience of the established parliamentary
democracies of western Europe. Our early analysis of what we called
1
This was mitigated by the weak cohesion of American parties, which meant that even if there
was little incentive for accommodation of the House Republicans as a party, the votes of
individual Republican members of the House (and Senate) frequently were required by the
majority party if it wanted to pass significant legislation.
Preface vii
“the three faces of party organization” (the party central office, the party in
public office, and the party on the ground—Katz and Mair 1993), although in
some important ways paralleling V. O. Key’s (1964: 163–5) conception of
American parties as comprised of “the party organization,” “the party in
government,” and “the party in the electorate,” assumed a more formal
structure, and particularly a more formal sense of party membership and a
more formal boundary between the party itself and a penumbra of loyalists
and supporters (both individuals and organizations) than found in the United
States. Nonetheless, even if our schema fits parties with formal membership
structures more directly than it fits those without, the underlying insight, that
all political parties—including those with only one member like Geert Wil-
ders’ Partij voor de Vrijheid or parties essentially paid for and run by a patron
like Silvio Berlusconi’s Forza Italia—should be understood as political sys-
tems in their own right remains. Moreover, we would suggest that the general
principles that we suggested shape competition and cooperation among the
three faces of membership-based party organizations should be expected to
apply, mutatis mutandis, to other types of party organizations as well.
Similarly, our historical/adaptive account of the evolution of party organ-
izations from elite to mass to catch-all to cartel initially appeared relevant
only to the countries of western Europe in which parliamentary institutions,
and thus rudimentary elite parties within parliament, developed before wide-
scale suffrage expansion, perhaps with the addition of the democracies of the
“old” British Commonwealth—and by an even greater stretch the addition of
the United States, which might be argued to have been an early example of the
catch-all model, but which never had approximated the mass party type. As
the idea of a party cartel as a way of accounting for contemporary political
events gained traction, however, it appeared to resonate with the experience of
countries outside of its locus of origin notwithstanding that they had not
experienced the same evolutionary processes. Moreover, although our ori-
ginal account of how a cartel party system evolved was rooted in the unique
historical experiences of western Europe, our account of the social, economic,
and political conditions that might lead to the establishment and maintenance
of such a system was not. Simply, it might be possible for a country to “skip”
some or all of the stages of the process and still arrive at the same result.
As with all theories dealing with complex social phenomena, it is impossible
to identify a crisp set of cases to which our hypotheses should be expected to
apply perfectly, and to contrast that to a crisp set of cases to which they should
not apply at all. Rather than trying to construct a dataset including all of the
variables, events, and processes in which we are interested for a well-defined
but comprehensive set of countries—a task that would in any case be
impossible—we have used, in addition to our own data, a variety of datasets
originally constructed by others to address other questions and then either
made publicly available in data archives or provided to us through the courtesy
viii Preface
of the original investigators. The selection of cases was decided by their
research priorities, with the result that our analyses are based on overlapping
but not entirely static sets of cases. In many cases, we have drawn on the work
of other researchers to provide examples without trying to replicate every
observation in the full range of cases; we can only leave it to the reader to
decide at what point a series of anecdotes cumulates to the status of data.
Any project that goes on for three decades accumulates an enormous
backlog of debts of gratitude to an enormous number of people and institu-
tions. Both constraints of space, and fear of inadvertently leaving someone
out, preclude attempting to name them all. Certainly, we are indebted to the
European Consortium for Political Research and its then chairman, Professor
Rudolf Wildenmann, for helping to launch the “party organization project,”
and to the NSF, the FGE, and the numerous other funding bodies that helped
to pay for it. None of this would have been possible without our collaborators
in that project. Ideas were tried out on generations of our students—some of
whom went on to do the research on which we have drawn in this book.
Numerous colleagues, friends, and conference participants have read and
commented on papers that later were incorporated into this work. Reviewers
from Oxford University Press made invaluable suggestions for improvement
to the completed draft. We have profited from their insights and are grateful
for their contributions, but also absolve them of any blame for what we have
made of their suggestions.
Finally, although this manuscript is being completed more than six years
after Peter Mair’s sudden and untimely death, it is indeed a co-authored work.
At the time of his passing, we had developed a full outline for the book, and
Peter had early drafts of three of the chapters for which we had agreed that he
would take the lead. While I have edited those drafts extensively—so that, as
I hope was the case with our earlier publications, it would not be evident
which of us had originally drafted what—his insights are reflected not only in
the chapters for which he wrote the first drafts, but in the chapters that I wrote
as well. This is his book as well as mine, although I am sure it is not as good as
it would have been had we been able to see it through to completion together.
One of the things I tell my students is that every book, no matter how
carefully researched and edited and read and proofread, inevitably will con-
tain mistakes. Notwithstanding what we say in the book about the desire to
politicians (like everyone else) to take credit and avoid blame, I accept that the
mistakes are mine.
Richard S. Katz
Contents
List of Figures xi
List of Tables xiii
1. The Problem 1
2. The Rise and Decline of Parties 29
3. The Locus of Power in Parties 53
4. Parties and One Another 81
5. Parties and the State 101
6. The Cartel Party 124
7. The Cartel Party and Populist Opposition 151
References 189
Index 209
List of Figures
1.1 Simple principal-agent model of democracy 3

1.2 Mass party principal-agent model with three social
segments/parties and coalition government 4
1.3 Downsian principal-agent model with three parties and
coalition government 5
2.1 The mass party and the catch-all party 50
3.1 Cotta and Best’s typology of legislators 79
6.1 Parties, civil society, and the state: the caucus party type 125
6.2 Parties, civil society, and the state: the mass party type 126
6.3 Parties, civil society, and the state: the catch-all party type 127
6.4 Parties, civil society, and the state: the cartel party type 127
7.1 Ties to groups and parties 161
List of Tables
3.1 Party leadership in Belgium, the Czech Republic, Poland,

Portugal, and Spain 59
3.2 Party membership as a percentage of the electorate, 1960–2010 63
3.3 Change in the numbers of party members 1980–2010 63
3.4 Incongruence of regional and national coalition before and
after 1999 71
3.5 MP annual base salary divided by median net household income 75
3.6 Parliamentary base salary 1976, 2012 76
4.1 Percentages of quasi sentences coded in “Labour Group:
Positive” for social democratic party manifestos and
“Labour Group: Negative” + “Middle Class and
Professional Groups: Positive” for liberal and conservative
party manifestos, 1950–70 and 1991–2005 85
4.2 Electoral volatility, 1945–65 and 1970–2004 86
4.3 Expert survey left-right party placements 88
4.4 Mean left/right positions of the left-most party in the social
democratic family and the right-most party in the liberal,
conservative, and Christian democratic families, 1950–70
and 1996–2005 89
4.5 Proportions of manifestos devoted to the economic cluster 92
4.6 “Governmental and administrative efficiency: positive,” “political
corruption: negative,” and “political authority: positive” 95
4.7 References to (party) government and (prime minister)
government in the Times (London) 1949, 1952, 1996, 2004 95
4.8 Numbers of new formulas, 1947–59, 1960–79, 1996–2015 98
4.9 Numbers of government formulas, 1947–69 and 1993–2015 98
5.1 UK appointments and reappointments to executive
non-departmental public bodies and National
Health Service bodies 122
6.1 Effective numbers of parties 133
6.2 Patterns of coalition formation, 1990–2015 135
6.3 Characteristics of party ideal types 141
xiv List of Tables

6.4 Percentage of British MPs whose “main prior Profession”
was “politician/political organizer” 142
7.1 Recent performance of populist parties in Europe 153
7.2 Effective numbers of parties in the first election after 2000 and
the last elections before 2017 155
7.3 Union density, 1982 and 2013 160
7.4 Percentage of the population that changed usual residence in the
preceding year 161
7.5 Average vote share of the mainstream parties by decade: Austria,
Germany, the Netherlands, and the UK 176
The Problem
There is little dispute with the idea that “democracy is a messy concept.”
Nonetheless, most political scientists, most democratic politicians, and most
of the growing “democracy-promoting industry,” share a common, and rela-
tively simple, understanding of democracy. At least in the modern age, they
agree with Joseph Schumpeter’s definition of democracy as a system “in which
individuals acquire the power to decide by means of a competitive struggle for
the people’s vote” (1962: 269). Moreover, in a large society meaningful
competition for the people’s vote requires both that the options among
which the people are asked to choose be sufficiently limited in number, and
that they be sufficiently coherent, that what might be called the “Ostrogorski
problem”1 can be mitigated. And providing those coherent options is identi-
fied as either a principal function, or else as the defining characteristic, of
political parties. Parties also are understood to provide the coordination
within representative assemblies, and across different branches or agencies
of government, that is required for the efficient conduct of business. As a
result, effective democracy is not just competition among individuals, but
competition among individuals organized into political parties. Both as def-
inition, and as the conclusion of an assumed causal process, democracy is
what results when people are free to form political parties, those parties
compete in periodic free and fair elections, and the winners of those elections
take effective control of the government until the next elections.
If there is little doubt that “democracy is a messy concept,” there is also a
growing consensus that “democracies are in a mess,” particularly with regard
to political parties. As we will show later in this book, parties have become
one of the least trusted political institutions; politicians are almost everywhere
the least trusted professionals; with a few upward blips, turnout in elections is
declining markedly, as is membership in political parties and identification
with them. If political parties are divided into two groups—the mainstream
parties that dominated post-war governments at least into the 1990s, on the
one hand, and populist or anti-party-system parties, on the other hand—
electoral support for the first group has declined (in many cases, plummeted
might be a more accurate description), while support for the latter has grown.
1
“[A]fter ‘the voice of the country had spoken,’ people did not know exactly what it had said”
(Ostrogorski 1903: vol. II, 618–19).
2 Democracy and the Cartelization of Political Parties

Not only have the post-1989 predictions of a universal triumph of liberal
democracy proven to be overoptimistic with regard to the former Soviet bloc
and the so-called Third World, even in its heartland of the first world the
future of liberal democracy appears less secure than only a few decades ago.
The natural question is how did this happen. E. E. Schattschneider’s often
quoted observation “that the political parties created democracy and that
modern democracy is unthinkable save in terms of the parties” and its less
often quoted continuation that “the condition of the parties is the best possible
evidence of the nature of any regime . . . The parties are not therefore merely
appendages of modern government; they are in the center of it, and play a
determinative and creative role in it” (1942: 1), is representative of the
centrality accorded to parties in modern empirical analyses of democracy.
Following Schattschneider’s lead, we look to the parties to provide some of the
answers to the question of how this, admittedly only the latest, “crisis of
democracy” came about. In particular, we argue that the mainstream parties
have formed, or at least have behaved in ways that could lead an outside
observer to believe that they have formed, what is in effect a cartel. This cartel-
like behavior has been driven by rational adaptation to social and political
changes, but it has also rendered the mainstream parties unable or unwilling
(often in the name of behaving “responsibly”) to address many problems that
confront their societies. This, in turn, has opened a space for challenges not just
to the parties in power at any moment, but to the whole idea of liberal party
democracy. While this is the particular theme of Chapter 7, the entire volume is
directed at laying the groundwork for that analysis.
In arguing for the centrality of political parties to any understanding of
democracy, Schattschneider (1942: 16) also complained that “the political
parties are still the orphans of political philosophy.” As van Biezen and
Saward (2008) say, that complaint remains largely true seventy-five years
after it was originally published. At a more mundane level, however, the
perceived centrality of parties has led to widely accepted, and in some cases
quite specific and detailed, prescriptions regarding how both parties and
government more generally should be organized. These prescriptions fre-
quently have been justified by a particular, albeit at the same time somewhat
vague, idea of democracy as “democratic party government”(Castles and
Wildenmann 1986; Katz 1987; Rose 1974). This, in turn, is often elaborated
in the increasingly popular terms of a “principal-agent” model of party
politics (Müller 2000; Strm et al. 2003).
This principal-agent model and its associated prescriptions for the organ-
ization and behavior of individual parties, and for the relationships among the
several parties, and among parties, citizens, and the state has exercised strong
influence over the way both social scientists and “political engineers” think
about establishing and maintaining healthy democracies. Our contention in
this book, however, is that this model in fact has only quite marginal
The Problem 3
connection to the ways in which parties and party systems really work in the
early twenty-first century. Moreover, we contend that the disconnect between
the normative justifications of, and prescription for, party democracy, on the
one hand, and the contemporary realities, on the other hand, is an important
contributor to the current malaise. Many of the empirical claims about parties
and party systems that we will be making—for example, that party member-
ship has been declining nearly everywhere—have been recognized for some
time. They have, however, generally been recognized only one at a time, and
interpreted as independent “problems” that can be addressed individually,
and rectified within the established principal-agent framework for under-
standing party government. In contrast, we propose a comprehensive frame-
work that explains how these individual findings hang together, how they
came about, and how, in particular, they undermine both the empirical
validity and the theoretical utility of the standard principal-agent model of
democracy—and how, in doing so, they pose an important challenge to the
survival of party government—and potentially to the survival of democratic
government as understood through the latter half of the twentieth century and
beyond more generally.
THE SIMPLE PRINCIPAL-AGENT MODEL

OF PARTY GOVERNMENT
In its simplest form, the principal-agent model of democracy in a parliamentary

system can be portrayed as illustrated in Figure 1.1. Starting on the right-hand
side of the figure, the apparatus of the state (particularly the bureaucracy)
works as the agent of the ministry, exercising authority delegated to it by the
Electorate
Party or parties in Ministry State

power administration
F I G U R E 1 . 1 Simple principal-agent model of democracy

Source: Katz (2014)

ministry in the pursuit of objectives set by the ministry. The ministry, which is
thus the immediate principal of the state apparatus, is simultaneously the agent
of the parliament, which is to say of the parliamentary majority as organized in
a coalition of parties. Finally, the parties in parliament act as the agents of the
electorate. The result is an unbroken chain of principal-agent links from the
electorate to all of the government (parliament, ministry, state apparatus),
ultimately making all of the government the agent of the electorate, and thereby
rendering the whole arrangement democratic.
This highly schematic rendering of democratic party government glosses
over many significant variations. Particularly from a European perspective,
attention primarily focused on variants of what Beer, drawing on the British
case, labeled “Socialist Democracy,” rooted in both the social and the polit-
ical theory of the mass party of integration (Beer 1969: ch. 3; Duverger 1959
[1951]: bk 1). The democratic theory associated with this can be expressed
as the principal-agent model illustrated in Figure 1.2, in which the single
“parties” box from Figure 1.1 is disaggregated into three separate parties, to
allow the idea of elections as competition among alternatives, and at the next
stage to allow the distinction between electoral winners and losers, to be made
explicit. In this version of democratic party government, each party is the
“political committee” of a particular segment of society (for example, of a
social class or confessional group) and acts as its agent, with the social
segments collectively encompassing the entire body of citizens. A coalition
of the parties in parliament then negotiates the formation of a ministry as their
agent; assuming that it is a majority coalition, and further assuming that its
majority in parliament reflects the support of a majority in the electorate, it is
therefore also the agent of the electoral majority—and if one accepts the
principle that the majority is entitled to decide/act for the whole, it becomes
the agent of the whole electorate. Finally, the ministry employs the state
Electorate Party 1
Segment 1 Ministry State

administration
Segment 2 Party 2
Segment 3 Party 3
F I G U R E 1 . 2 Mass party principal-agent model with three social segments/parties

and coalition government
Source: Katz (2014)
The Problem 5
apparatus as its agent. The direct principal-agent chain from voters to parties
to ministry to administration summarized in Figure 1.1 is thus maintained,
with the administration still the ultimate agent of the voters.
Particularly in the later decades of the last century, an alternative version of
this model, derived from economic theory and identified eponymously with
Anthony Downs (Downs 1957), rather than being derived from sociology as
interpreted, for example, by Lipset and Rokkan (1967), came to prominence.
In this model, parties are teams of politicians (Downs 1957: 25; Schumpeter
1962 [1942]: 283; Schlesinger 1994: 6), rather than associations of citizens, and
compete to be “hired” as the agents of the whole society, rather than operating
as the already established agents of particular social segments. The principal-
agent understanding of democracy, at least in stylized form, however, appears
to be virtually the same—especially if the primary competitors are assumed to
be either two parties or two distinct and stable coalitions. Even in a multiparty
case, the graphic representation in Figure 1.3 appears essentially the same as
that illustrated in Figure 1.2. The voters as principals choose a party to act as
their agent, although in this case it is not majority support for a particular
party or coalition, but rather that the governing coalition includes the party
that represents the first preference of the median voter, that underpins legit-
imacy, whether or not the cabinet represents a majority coalition. The party
(or coalition of parties) in parliament installs a ministry to act as its agent. The
ministry employs the state apparatus as its agent. Yet again, government is the
ultimate agent of the voters, and the system is, therefore, democratic.
This model (at this level of generality, it is reasonable—and common—to
regard the models in Figures 1.2 and 1.3 simply as variants of the simple
model in Figure 1.1) is very comforting for those who would like to reconcile
the realities of modern politics with a normatively informed vision of democ-
racy as “government of the people, by the people, for the people,” although its
Electorate Party 1
Ministry State
administration
Party 2
Party 3
F I G U R E 1 . 3 Downsian principal-agent model with three parties and coalition government

Source: Katz (2014)

appropriateness as an empirical model, as opposed to a normative ideal, has
always been open to some question. Indeed, Blyth and Katz (2005) have gone
so far as to suggest that the entire model might be reversed, with the cabinet
acting as the agent of the administration (individual ministers arguing for
their department’s policies and budget) rather than its principal, the cabinet
buying the support of MPs with the prospect of career advancement and
promotion of pet policies, and the parties buying voters with policy promises
and patronage.
Be that as it may, like all principal-agent relationships, those portrayed in this
model are subject to “agency slack.” One of the “core assumptions” of the
canonical principal-agent is that the interests or preferences of the agent differ
from those of the principal (Miller 2005: 205–6), and this creates incentives for
shirking by the agent. Much of the literature on principal-agent relationships
concerns ways in which such shirking can be contained, but at its base, it still
retains the basic idea that initiative lies with the principal, so that outcomes
ultimately can be traced to the interests or preferences of the principal. From this
perspective, agency slack accounts for observed failures of agents to act opti-
mally in the interests of their principals in much the way that friction accounts
for the failure of falling objects to conform exactly to the predictions of the
simple equations of first-year physics. As some exponents of “behavioral eco-
nomics” (e.g., Cartwright 2011; Diamond and Vartiainen 2007) have argued in
contrasting their approach to that of classical (or “rational choice”) economics,
it is possible for the divergences between model and reality to become so great
that the model no longer provides even a useful baseline against which diver-
gences can be assessed, and becomes instead an impediment to understanding.
For the principal-agent model to be appropriate for describing the relation-
ship between citizens and parties requires that ultimate power rests with the
citizens as principal. In Sappington’s (1991: 47) words, “The principal is
endowed with all of the bargaining power . . . and thus can make a ‘take-
it-or-leave-it’ offer to the agent.” As translated into the electoral sphere, this
means at least that, on the one hand, the electorate must have a substantial
choice among competing parties, and, on the other hand, that the cost of the
potential sanction of electoral defeat to a party is sufficiently high as to
“concentrate the mind wonderfully.” The essence of our argument, first
advanced some twenty years ago (Katz and Mair 1992a, 1995) but even
more true today, is that these conditions are not well met in modern democ-
racies: the choice offered to electors by the “mainstream parties” (i.e., those
with a realistic chance of being in government in the medium term) has
become progressively less substantial in the sense that changes of government
are less directly tied to changes in policy or outcomes, and the cost to parties
in the mainstream of losing an election (the difference in pay-offs between
being a winner and being a loser) has been significantly reduced. Going
beyond this simple observation, we make two additional claims. On the one
The Problem 7
hand, even if these changes can in some ways be traced back to long term
social processes, many of these social processes are, in their turn, the result of
government policies, and thus they are only partially exogenous to the parties.
On the other hand, it is most immediately the intentional responses of the
parties to these social processes, not the social changes themselves, that have
undercut the basis for a principal-agent understanding of party government.
In particular, our argument is that at the level of party systems, the main-
stream parties, and most minor parties as well, have effectively formed a
cartel, through which they protect their own interests in ways that sap the
capacity of their erstwhile principal—the electorate—actually to control the
parties that are supposed to be the agents of the electorate. While the appear-
ance of competition is preserved, in terms of political substance it has become
spectacle—a show for the audience of “audience democracy” (Manin 1997; de
Beus 2011). Further, we argue, in order to facilitate this cartel-like behavior,
political parties have adapted their own structures, giving rise to a new type of
party organization, which we identify as the “cartel party.”
This book is devoted to connecting these twin developments of waning
substantive competition and political party transformation, along with the
social, historical, and political processes that underpin them, to understanding
their impact on both the practice of, and popular support (or not) for,
democratic government, and to considering what these processes mean for
the future of liberal democratic party government.
PARTY CHANGE
As is true of virtually all social processes, with the benefit of hindsight the
roots of these developments can be found reaching back well before they were
generally recognized to be significant—in our case, at least to the 1950s. Also,
like most general social processes, they developed at different times and at
different rates (and from different starting points) in different countries. Their
acceleration and confluence at a level sufficient to pose a serious challenge to
the practices and legitimacy of established institutions of party government
are of fairly recent origin, however. We do not suggest that there was some
golden age in which democratic party government functioned smoothly and
with unquestioned legitimacy. Nonetheless, while the party government
model was always an ideal type rather than a fully accurate description, an
array of social changes have occurred, accompanied by changes in the parties
themselves, that have moved reality so far away from the ideal type that even
its heuristic utility must be questioned. The result is a far less sanguine view

than the “triumph of democracy” literature (e.g., Mitchell 1997; Preston 1986)
might lead one to expect.
At least into the 1980s, most theory and research concerning political
parties, at least outside of the United States, was premised on the assumption
that the norm, both empirically and evaluatively, was either the mass party of
integration, or else the more modern catch-all party, still understood to be a
variant of the mass party. This was what parties in democratic polities should
be like, and how they should be organized and behave. To the extent that they
did not meet these standards, they were, essentially by definition, somehow
weak or failing. Philippe Schmitter’s (2001) critical evaluation of the role of
parties in the consolidation of the new democracies of the last quarter of the
twentieth century provides a good example of the persistence of this mode of
thinking. Even in the 1990s, however, it was apparent to some observers that
the process of party organizational development and adaptation was more
varied, more fluid, and more open-ended than that narrow conception
allowed (Katz and Mair 1994).
In particular, the decline in partisan attachments (party identification, party
membership, electoral turnout), declining social segmentation, increasing
education and leisure time, all appeared to be undercutting the assumptions
upon which the mass party model had been constructed. Simultaneously, the
economic model upon which many government policies, especially those that
defined the welfare state, had been built was also being called into question.
Not surprisingly, accounts of party change (e.g., Katz and Mair 1992a)
focused almost exclusively on domestic factors, whether social, political,
economic, or institutional.
In retrospect, it is clear that the influence of factors drawn from the world of
international politics might have been taken into account even then, and
certainly need to be included now. In the early 1990s, economic globalization
began to be recognized as a serious constraint on the capacity of all govern-
ments to manage their national economies. In 1989, the Berlin Wall fell and
the Soviet Bloc, and then the Soviet Union itself, began to crumble. In
February 1992, the Maastricht Treaty was signed by the member states of
the European Union (EU), with the national currencies of all of the then
members of the EU except the United Kingdom (UK), Denmark, Sweden,
and Greece (which joined the rest in 2001) replaced by the euro on January 1,
1999, ending national control over monetary policy in the eurozone countries.
In January 1995 the World Trade Organization (WTO) was established. The
events between 1989 and 1999 obviously brought about major changes in
international affairs, but they also had a profound impact on domestic politics
in the advanced industrial democracies. Although Maastricht, the treaties
that followed it, and the introduction of the euro are specific to the EU, the
impact of the collapse of the Soviet empire and of economic globalization and
the WTO has been felt far more widely.
The Problem 9
These developments substantially undermined the stakes of traditional
electoral competition, first by reducing the perceived importance of the left-
right ideological divide that lay at the heart of most Western party systems,
and that, whether implicitly or explicitly, fed off the Cold War divide; second
by transferring control over a range of economic (and other) concerns beyond
national borders to technocratic and largely non-partisan institutions like the
EU system, the WTO, the International Monetary Fund, and World Bank—
and to multinational corporations, some of which have budgets larger than
the GDPs of many of the countries in which they operate; and third, even
beyond the formal transfer of powers and responsibilities to institutions like
the EU or the WTO, by facilitating an ideational shift (Blyth 2002) suggesting
that what had traditionally been the central political concerns of inflation and
unemployment now properly lay outside the control of national governments,
and thus were no longer among the core responsibilities of the parties that
formed those governments. We address all of these issues elsewhere in this
volume, and particularly in Chapter 4.
These changes in the international arena interacted with the tendencies
already noted in the domestic arena to give all significant political parties, no
matter how bitter their rivalries had been in the past—and indeed no matter
how intense their rivalries might appear to be in the present—a core set of
common interests and common constraints, and thus also common incentives
to cooperate, and to collude, to protect those interests. Cooperation and
collusion, which are obviously important elements in our cartel thesis, become
easier when the stakes of competition are reduced, and this was one of the
results of the shedding of responsibility for managing the economy and of the
end of the existential struggle between the “free” and “communist” worlds.2
THE CARTEL THESIS
We initially arrived at the idea that new patterns of relationships were

emerging among parties, society, and the state, among the parties themselves,
and within individual parties among their various “faces” (Katz and Mair
1993) inductively from a data-gathering project whose primary purpose was
to document changes in party organizations from 1960, when the mass party
was widely believed to be losing ground to the catch-all party as the dominant
2
While Huntington’s (1966) struggle between Muslim and Western worlds may have an
equivalent existential import, it does not represent a cleavage within the Western democracies
with which we are concerned, because unlike the cleavage between socialism and capitalism, there
have been no significant Islamist parties in the Western democracies.

form of party organization in Europe. Although the project did not start from
any particular hypothesis or theory, our attention was quickly drawn to a
series of real-world developments that appeared to be striking, to be reason-
ably pervasive, and not to have been widely noticed or studied. Two of these
in particular need to be emphasized.
The first was the evident transformation of parties from purely private
organizations, structured by their own constitutions, governed by their own
rules and procedures, and funded by their own resources, into organizations
that were ever more controlled by regulations laid down by the state and ever
more dependent on resources provided by the state. Although Kirchheimer
had already noted in the 1950s (e.g., Kirchheimer 1957, see also Krouwel
2003, 2006: 258–60) that parties were being drawn into what he saw as an
excessively close relationship with the state, most work on party organiza-
tional change—including that of Kirchheimer himself—focused on the society
as the driving force and as the place where explanations for party change
could best be sought (for an exception, see Müller 1993).
In contrast to society-driven explanations of party change, we saw decisions
made by the state and embodied in law to be of great significance. One of these
decisions that stood out in particular in our early work was the increasingly
widespread practice of providing the parties with substantial state subventions
to fund party organizations both within and outside of parliament. These
subventions were often accompanied by party laws that laid down, sometimes
in quite detailed terms, what parties could or could not do, not only with
regard to the use of these state-supplied funds, but with regard to privately
raised funds, and indeed with regard to their organizational practices more
generally. Access to public service broadcasting and sometimes even commer-
cial broadcasting and media, which were becoming more and more important
for party campaigning and publicity, was increasingly a subject for detailed
state regulation. Simply, parties were becoming less able to make their own
decisions without reference to legal restrictions—and because those restric-
tions applied to all parties, there was less room for parties to distinguish
themselves from one another in organizational terms.
If, as now seems undeniable, parties are strongly influenced by the state,
and indeed in a real sense are drawing closer to, and more involved with, the
state, might they also be drawing further away from society? In our original
papers, we suggested that this was the case, without presenting systematic
evidence. Later research, summarized very comprehensively in Dalton and
Wattenberg (2000; see also Mair 2013), has suggested that the ties between
parties and society are indeed becoming more tenuous: there has been a sharp
decline in party membership in the 1990s and into the 2000s; there has been a
consistent decline in levels of party identification; there has been a somewhat
more erratic but nonetheless pronounced fall in voter turnout. We discuss this
later in the volume, and in particular in Chapter 2.
The Problem 11
Further, if parties are increasingly influenced by the state, and in particular
by regulations that apply to all parties, then it is likely that they will also come
increasingly to resemble one another. Many things would be shared by all
parties, including their means of communication, their principal sources of
finance, their internal organizational form and modes of adapting to party
laws, and their ever more common experience of holding public office—see
especially Chapters 4 and 5. In other words, when speaking of party experi-
ences or the nature of a party, it had begun to make more sense to speak in
terms of “the parties” or “the party system” rather than in terms of any
individual party or party “family.” To be sure, the influence of the state on
the parties was only one of a number of factors pushing parties to resemble one
another and thereby promoting organizational convergence (Epstein 1967).
Other influences stemmed from social changes that led the parties to appeal to
similar and overlapping constituencies and from the development of modern
campaign technologies. Adaptation to party laws, state subvention require-
ments, and the exigencies of holding government office were also crucial,
however, and these factors had often been overlooked by the literature.
Moreover, although parties were more influenced by the state, by public
regulations, and so on, this did not imply that they were being influenced by
something that was entirely exogenous to themselves. The laws and rules
influencing parties were those that they themselves, as governors, had been
centrally involved in writing. Indeed, the parties are unique in that they have
the ability to devise their own legal (and not only legal) environment and,
effectively, to write their own salary checks. As van Beyme (1996: 149)
observed, “the new political class as a transfer class was privileged in two
respects: by being the only elite sector which determines its own income, and
by organizing state-support for the organizations which carried them to
power, e.g., the parties.”3
Given all this, it also makes sense for us to expect that parties would
cooperate with one another. In fact it is generally necessary (or at least
politically expedient) for parties to cooperate with one another if general
party regulations are to be written and if a system of public financing is to
be introduced. And it is clearly a small step from consideration of cooperation
and agreement, particularly with regard to measures perceived by the parties
to be necessary but unpopular like increasing subsidies for themselves, to
consideration of collusion. But to recall: all of this starts from the empirical
3
The claim that the political class is the only elite sector that determines its own income is
probably a bit exaggerated, as the compensation packages of corporate CEOs, often determined
by “compensation committees” made up of the CEOs of other corporations, illustrate. And both
have led to complaints of self-serving behavior in which the interests of constituents (voters in the
case of politicians; shareholders in the case of CEOs) are sacrificed to benefit those making the
decisions.

observation that shows that parties are moving towards the state. And while
what follows in our reasoning may be contested or qualified, the original
observation nevertheless still stands.
From the very beginning we conceived of party organizations as being akin
to political systems, with three important constituent units or “faces” (see also
Chapter 3). This was different from the approach commonly used up to that
point to discuss party organizations, which simply distinguished the party in
parliament and the extraparliamentary party. Our division was similar to, but
not identical with, V. O. Key’s (1964: 164) tripartite division of US political
parties into the party as organization, the party in government, and the party
in the electorate. Our concern was only with the party itself, and for the most
part with parties that (unlike American parties) have formal membership
organizations, and hence we distinguished among the party in public office
(PPO), which included the party both in parliament and in government; the
party in central office, which was constituted by the permanent bureaucracy,
national executive organs, and so on; and the party on the ground—the
organized membership. We had expected that the balance among these
might shift, and this is indeed what we found. This led to our second uncon-
tested finding: in those long-established democracies for which we gathered
data, the weight of power within the party, as measured by changes in the
locus of decision making, as well as by the distribution of internal resources—
finance, staff, etc.—has moved much more firmly into the hands of the party
in public office.
This finding then led to additional hypotheses that subsequently fed into the
general cartel thesis and which, of course, proved more disputable. The first of
these emphasized the sheer self-interest of those actors who actually occupy
the public offices in the name of the parties and who, like the politicians and
administrators observed by Skocpol (1992: 40), “have ideas and organiza-
tional and career interests of their own, and they devise and work for policies
that will further those ideas and interests, or at least not harm them.” Our
hypothesis was and is simply this: that as the party in public office gains
ascendancy within the party as a whole, its particular interests will be treated
as being the interests of the party writ large. We discuss this at greater length
in Chapter 3. Moreover, although it might seem at first sight that the interests
of the PPO could be summarized simply to lie in winning, in our view it made
more sense to see those interests as lying equally in having the possible costs of
losing reduced as much as possible. After all, always winning is unlikely,
either for parties as organizations or for many of their candidates as individ-
uals.4 We also further hypothesized that this would be true for the PPOs in all
(mainstream) parties. And this, in turn, would be likely to encourage a system
4
The exceptions include candidates nominated for safe seats or to the top of closed PR lists.
The Problem 13
of mutual cooperation that should, under normal circumstances, lead to the
emergence of a Nash equilibrium: an equilibrium or compromise from which
no one participant will have an incentive to defect.
Putting these two sets of findings and their related hypotheses together leads
to the following conclusions. First, parties are increasingly part of the state,
and increasingly removed from society, and this new situation encourages
them, or even forces them, to cooperate with one another. They can write their
own checks, but only if there is general agreement to do so. Second, these
parties increasingly resemble one another; in terms of their electorates, pol-
icies, goals, styles, there is less and less dividing them—their interests are now
much more shared, and this also facilitates cooperation. A very important
part of their shared interest is to contain the costs of losing, and in this sense to
find an equilibrium that suits all of their own “private” interests. This also
means cooperation, even if this cooperation need not be overt or conscious.
That is, even if parties might be disinclined to rely heavily on overt deals with
one another, their mutual awareness of shared interests, and their sense of all
being in the same boat and relying on the same sorts of resources, means that
we can conclude by hypothesizing collusion (or its functional equivalent) and
cartel-like behavior.
CARTEL PARTIES AND A PARTY CARTEL
Although the idea of a cartel implies concerted action, when translated into
the cartel party model the term was not intended to imply or depend on an
actual conspiracy and it is particularly in this respect that the choice of
denomination may have been less than perfect (Chapter 6). Rather, as anyone
involved with regulations or legislation concerning anti-competitive practices
in the economy is well aware, it is possible to produce the effects of collusion
without any illicit communication or covert coordination (e.g., Werden 2004).
In an oligopolistic market, which the electoral market with only a handful of
parties receiving nearly all of the votes certainly approximates, overt signaling
can produce virtually the same result as covert conspiracy.
The denomination “cartel” also implies attention to interparty or system-
level dynamics, and in particular to a distinction between those players that
are “within” the cartel and those that are excluded from it. Indeed, part of
the original argument was that participation in a cartel-like pattern of con-
strained competition with other parties would both facilitate and, at least to a
certain extent, require many of the changes in internal party arrangements
that we identified with the cartel party as an organizational form. Thus
even if analytically separable, the idea of a party cartel as a system-level

characteristic and the idea of a cartel party as a type analogous to the mass
party or the catch-all party are closely intertwined.
Attention to the system-level or interparty side of the argument requires
that a further point be clarified, and that is the specification of the set of
parties that are expected to be “in” the cartel. We have sometimes identified
this as the set of “governing” parties (Katz 2002, 2003). Unfortunately in
practice this phrase has proven to be slightly ambiguous, but what it clearly is
not intended to denote is simply those parties that are in government (holding
ministerial portfolios, or the equivalent) at any particular time. While it does
not necessarily extend to all parties that might in theory be considered as
potential coalition partners (i.e., that are not excluded from government on a
priori grounds)—indeed, one of the hypothesized characteristics of a cartel
system is to minimize the importance of the distinction between being in and
being out of office at any particular time—or that play a governing role in any
subnational government, it does extend to all parties that have a reasonable
expectation that they might be included in a national governing coalition or in
a significant share (defined jointly by number, size, and range of competences)
of subnational governments within the reasonably foreseeable future. More-
over, while a cartel does imply constrained competition, this refers to the
nature of the competition rather than to an absence of electoral turnover—to
the question of whether it makes any difference who wins, not to the frequency
with which different parties win. Indeed, the absence of an expectation of
turnover would be a factor strongly militating against the formation of a
cartel. Thus, that the American Republicans in the House of Representatives
appeared in the early 1990s to be condemned to permanent opposition
status was a major contributor to Newt Gingrich’s “Contract with America”
as a way to fracture a cartel that arguably included both Democrats and
Republicans in the Senate and in presidential politics.
MAKING SENSE OF CARTELIZATION
Given this background and these clarifications, our argument can be sum-
marized relatively briefly. At least by the 1970s, the dominant form of party
organization in most democratic countries approximated what Kirchheimer
(1966) had identified as the catch-all party. While there were still obvious
connections, both in terms of formal organization and affective ties, between
particular parties and particular social groupings, these had noticeably weak-
ened. Increasingly, parties were seen, and saw themselves, as brokers
among social groups and between social groups and the state, rather than as
the political arms of specific groups. Ideological conflicts and deep social
The Problem 15
cleavages had been transformed into amorphous differences in general
left-right orientation. A significant component of electoral competition
involved the provision of public services, with parties in effect bidding for
support from voters by promising more services (especially on the left) and
lower taxes (especially on the right), and for support from potential contribu-
tors by offering specially tailored legislation that often resulted in subsidies to
special interests, the weakening of otherwise desirable regulation, or the
collection of less revenue.
This situation confronted the parties with three interrelated classes of
problems, some of which might be characterized as largely exogenous, but
others of which were largely the result of actions taken by the parties them-
selves in the past. First, the moderation of class and other subcultural con-
flicts, and the increasing homogeneity of experiences and expectations of the
vast majority of citizens associated with the rise of mass society and the
welfare state (mass media and mass culture, mass education, near universal
provision for health care, unemployment, and old age insurance) reduced the
value of appeals to class or cultural solidarity. Concurrently, the process
identified by Inglehart (1970, 1990) and Dalton (1984) as “cognitive mobil-
ization” contributed to a general decline in affective attachment to parties per
se as part of a process of partisan dealignment. Not only party psychological
identification, but formal party membership, declined. As the other side of the
same coin, electoral supporters (party members, party voters, organizational
contributors) became less reliable.
Second, with the increasing reliance on mass media as the most effective
mode of campaigning, and with the attendant increase in the need for profes-
sional expertise (pollsters, advertising consultants, direct-mail fundraisers and
marketers), the economic costs of remaining competitive were rising more
rapidly than the ability or willingness to pay on the part of the party on the
ground. The initial response of turning to a range of interest organizations
(primarily unions) and corporations also began to reach the limits of willing-
ness to pay, at least without quid pro quos bordering on, or entering, the
realm of the corrupt. These changes also meant that the non-monetary
resources that the party on the ground could bring to the table (e.g., volunteer
labor for campaigning; knowledge of local opinion) were becoming relatively
less valuable to the party in public office (in comparison to mass media space
or information gathered by professional pollsters).
Third, at least if one accepts the idea that there is a real limit beyond which
the provision of public goods cannot be expanded without creating a fiscal
crisis, then the governments of many welfare states appeared to have backed
themselves into a corner from which the only escape without, and potentially
even with, untenable tax increases was equally untenable service cuts. More-
over, servicing the public debts that accumulated while deferring addressing
this dilemma ultimately made even that “strategy” increasingly untenable.

Although of a different type, one additional development can be added to
this list. As politics has become an increasingly specialized profession, the
potential personal costs for politicians and party functionaries of electoral
defeat or organizational contraction have increased. Further, the separation
of parties from ancillary and other interest organizations that was character-
istic of the catch-all party has proceeded even further, and has reduced the
availability of jobs in those organizations for politicians who are (to use the
theatrical euphemism) “resting” between engagements. This loss of “out-of-
office” employment possibilities has only been partially mitigated—and
mitigated only for particular types of politicians—by the possibility of post-
politics careers in journalism or lobbying and, at the very top, making public
speeches for large fees. Simply put, when politics is a person’s primary
source of income, the stakes are higher (Borchert 2000; see also Chapter 3).
One implication of this is to reorient the meaning of party rationality
away from maximizing the expected (average) pay-off or probability of
victory, and toward maximizing the reasonably anticipated minimum pay-
off (“maximin”) even in defeat. Significantly, this is something that all the
mainstream parties can do simultaneously.
These problems are shared by all governing and would-be governing par-
ties, and set up the conditions for the formation of what is effectively a cartel,
in which participating parties serve their joint interest in providing for their
own security and survival. In terms of relations among parties, this has two
primary aspects. The first is restriction of policy competition, with policy
promises effectively playing the role of quantity offers in an economic cartel.
This is evident in the increasingly common moves to take issues out of the
realm of party competition by delegating them to non-partisan agencies like
independent central banks, courts, or the EU Commission, by privatizing
previously public functions (e.g., pension reform or health-care reform), and
by the increasingly common acceptance of various models of governance, new
public management (Hood 1991), and the regulatory state (Majone 1994,
1997), all of which privilege questions of technical and managerial expertise
over those of values or political preference (see Chapter 4). Even in the case of
issues that have not explicitly been removed from the realm of partisan
debate, cartel parties limit the degree to which they attempt to “out-bid”
one another. The result is that many issues are simply avoided by the main-
stream parties as demagogic or populist, and the range of proposals offered
for those issues that remain is often limited in the name of “realism” or
“responsibility.”
The second aspect involves attempting to solve the problem that internally
generated funds prove inadequate to the exigencies of modern politics, and to
mitigate the risks of electoral misfortune by reducing the disparity of
resources available to those in and out of government at any particular
moment, in both respects by turning to the coffers of the state. In the first
The Problem 17
respect, state subventions become significant—in some cases helping to fill the
gap between traditional sources of party income and perceived needs, and in
others largely replacing private contributions. In the second respect, a system
in which the parties of the ruling coalition enjoyed the resources of the state
(the power to appoint to office (and perhaps to “tax” the appointees), the
research capacities of the civil service, etc.) while the other parties were left to
their own devices is supplanted by arrangements that allow all of the cartel
parties to share in the bounty, and thus to reduce the pecuniary difference
between being in office and out of office.
CARTEL PARTIES
Cartels face two potential threats. One, as Kitschelt (2000) has pointed out, is
defection. The other is challenge from new entrants. Thus an additional aspect
of the cartel is the structuring of institutions such as the financial subvention
regime, ballot access requirements, and media access in ways that disadvan-
tage challengers from outside (Bischoff 2005). Moreover, because parties are
not unitary actors, the leaders of the party in public office (from whose
perspective this model has been developed) face not only the threat of defec-
tion or challenge by new party entrants, but also pressures or threats from
within their own party. It is in responding to these challenges that parties tend
to become cartel parties with respect to their internal structures.
One aspect of this has already been mentioned: by turning to state subven-
tions, parties—that is, their leaders—become less dependent on members and
other contributors.
A second aspect is the disempowering of the activists in the party on the
ground, who are the ones most likely to make policy demands inconsistent
with the “restraint of trade” in policy that is implied by the cartel model.
Although the objective is a kind of party oligarchy, the means ironically (or
not, depending on one’s reading of Michels (1962 [1911]) and the “iron law of
oligarchy”) may be the apparent democratization of the party through the
introduction of such devices as postal ballots or mass membership meetings at
which large numbers of marginally committed members or supporters—with
their silence, their lack of capacity for prior independent (of the leadership)
organization, and their tendency to be oriented more toward particular
leaders rather than to underlying policies—can be expected to drown out
the activists.
A third aspect is the centralization and professionalization of the party
central office (in particular, emphasizing the cash nexus of an employment
contract instead of partisan loyalty or ideology as the basis for commitment),

or ultimately even the outsourcing of campaigning and of the other functions
of the central office and the traditional party on the ground, again with the
result of freeing the leadership of constraints from below.
One consequence of all this is that while it may be appropriate to attribute
functions to parties, for example to provide a linkage between citizens or
social groups and the state, within the context of a theory about how demo-
cratic governments should work, it is not necessarily appropriate to assume
that parties (or more accurately their leaders) do give the performance of these
functions the highest, or even high, priority over such other potential goals as
personal power or economic/job security. In particular with regard to party
finance, the claim is not that state subvention makes it more difficult for
parties to provide this linkage (e.g., “extensive reliance on the state for
funding contributes to an erosion of parties’ capacity to link society and the
state”—Young et al. 2005), but rather that it reduces the parties’ need or
desire to do so, and thus is likely to reduce the degree to which parties actually
provide linkage, even if their hypothetical capacity to do so were increased by
access to additional funds.
The cartel party model also further cements the relationship between
parties and the state. With significant policy competition largely precluded,
whether as part of cartelization, or because of domestic fiscal and political
constraints, or because of the ever more powerful international constraints,
party spokesmen tend to become apologists for and defenders of policies that
they have inherited from their predecessors (Rose and Davies 1994) or, more
recently, have been imposed from outside, for example by the European
Commission (“Brussels made me do it!”—Smith 1997) or by the “troika” of
the Commission, the European Central Bank, and the International Monet-
ary Fund, and that have thus become more generically policies of the state
than they are the policies of any particular party or coalition. Moreover, as
part of the price for state funding, parties have also accepted that there will be
a significant body of regulations limiting both their activities and their struc-
tures, regulations which they themselves then developed. In this way, parties
move beyond the public utility model of regulation discussed by van Biezen
(2004; see also Epstein 1986) to approximate, in effect, full-grown institutions
of the state.
CONSTRAINTS ON CARTELIZATION
It is important to emphasize that the cartel party remains an ideal type, which
may be approximated or approached but which will not be fully realized—just
as there never were any parties that fully met the ideal type definitions of the
The Problem 19
mass party or the catch-all party (Katz 2017). Even with that said, however,
two forces restraining the cartelization of parties must be recognized.
The first restraint is that although the process of cartelization may be seen
as anti-democratic, parties, even in the cartel model—or perhaps particularly
in the cartel model—justify their own existence and their claim on state
resources on the basis of their contribution to democracy, and it is in this
respect that they are often open to challenge. On the one hand, cartelization
has clearly contributed to the rise of populist anti-party-system parties that
appeal directly to public perceptions that the mainstream parties are indiffer-
ent to the desires of ordinary citizens. Such parties have grown substantially in
both prominence and support in the last decade, and serve to underline the
dangers to cartel parties of excessive, or excessively overt, cartelization (see
Chapter 7). On the other hand, cartel parties also have to be attentive to the
potential backlash of being perceived to have excessively violated norms of
democratic fairness. While one would expect a certain level of disingenuous
rhetoric attempting to justify regulations that are in the parties’ interest as
actually being in the public interest, particularly with an aggressive free press
there will be real limits to the degree to which parties can construct institu-
tional biases in their favor without incurring even greater political costs.5
A second restraining factor is that although parties through their parlia-
mentary majorities make the rules that govern their own behavior and struc-
tures, govern entry to the political marketplace, and allocate state resources,
they do not do so with complete autonomy. Most obviously, and only
exacerbated by the increased role of courts, they are bound by constitutional
restrictions. Thus, although the basic logic of a cartel might lead one to expect
the ruling parties to restrict access to public finance to themselves (as to a great
extent they have done in American presidential elections6), German parties
were forced by the Bundesverfassungsgericht to provide public funding not
just to parties that clear the 5 percent threshold for representation in the
Bundestag, but to all parties that achieve one tenth of that result. Similarly,
5
With specific regard to reforming electoral laws to advantage those writing the reforms, see
Katz (2005).
6
“Major” parties, defined as those that received at least 25 percent of the vote in the previous
presidential election, are eligible for a subsidy; “minor” parties (those that received between 5 and
25 percent in the previous election) can receive a proportionately reduced subsidy; new parties or
those that received less than 5 percent of the vote in the last election can receive a similarly
proportionate subsidy—but only if they clear the 5 percent threshold in the current election, and
only after the fact. In 2000 (the last time a party other than the Democrats and Republicans
received a general election campaign grant), the campaign of Reform Party candidate Patrick
Buchanan received $12,613,452—in contrast to the $67,560,000 received by each of the major
party campaigns. In 1996, Ross Perot received $29,000,000 (the major parties each received
$61,820,000). Because acceptance of the general election campaign grants requires acceptance
of overall limits, the last major party candidate to accept the grant was John McCain in 2008.

in Figueroa v Canada (Attorney General) the Supreme Court of Canada over-
turned the provision of the Canada Elections Act that required a party to have
candidates in at least fifty ridings in order to reap the benefits of party regis-
tration, a requirement that would either have denied those benefits to most
small parties or forced them to bear the burden of nominating candidates in
many ridings that they did not intend seriously to contest.7
CRITICISMS
Although the cartel party thesis has become an important point of reference
for studies of political parties, it has not been without criticism. Ruud Koole,
one of the original collaborators in our data-collection project, and later
(2001–5) chairman of the Dutch Labor Party, raised a number of significant
points (Koole 1996), to which we responded at the time (Katz and Mair 1996);
the substance of those responses is reflected in the chapters that follow.
Perhaps the most elaborate critique was advanced by Herbert Kitschelt
(2000). He raises three basic objections, to which we respond briefly here
(see also Blyth and Katz 2005), although our real purpose now is to use
Kitschelt’s critique to call attention to basic disjunctures between our argu-
ment and the way it has been interpreted by some of its critics.
Kitschelt’s first complaint is with our claim (put in the terms of principal-
agent models, as exemplified by Figures 1.1 through 1.3) that parties and their
leaders have become less faithful agents of their electoral principals. He asks,
for example (p. 155), “[w]hy do parties wish to abandon their voters’ prefer-
ences . . . Would not vote- and office-seeking politicians attempt to realize
their goals by being more responsive to a greater share of the electorate
than their competitors?”8 But this, along with his doubts about the “state”
7
One of the costs was a requirement that each candidate post a deposit of CAD$1000. Prior to
2000, CAD$500 would be refunded only if the candidate received at least 15 percent of the vote;
after 2000, the full deposit would be returned upon satisfaction of reporting requirements, but a
small party might still be forced to borrow (presumably at interest) much of the $50,000 required
for fifty candidates.
8
Another complaint (p. 158) is that our “hypothesis asserting the empowerment of (generally
passive) members at the expense of local party activists is inconsistent with their claim that even
contemporary parties value activists and therefore permit greater participation in strategic
decision making.” But while we would not deny the utility of active members both as a source
of “free” labor and for increasing the apparent democratic legitimacy of the party, our suggestion
is that participation is broadened precisely to dilute the influence of activists, and thus to render
the leadership more, rather than less, independent in strategic decision making.
The Problem 21
as an alternative principal, means accepting the principal-agent model as
appropriate in the first place—contrary to our observations above. Moreover,
even if one were to accept in part the principal-agent framework for its
heuristic value, the answer to Kitschelt’s question just quoted would be that
one cause of cartelization is the desire of professional politicians to lessen the
force of the electoral incentive—making vote and office seeking less important
to the realization of their goals.
Kitschelt’s second complaint (p. 149) is that “inter-party cooperation gen-
erates a prisoner’s dilemma in the competitive arena that ultimately prevents
the emergence of cartels. Ideological convergence of rival parties has causes
external to the competitive arena, not internal to it.” This actually comprises
two claims: that cartels will not form, and that the causes of policy conver-
gence are exogenous to party politics. With regard to the latter, we appear to
disagree with regard to the meaning of exogeneity, our position being that
many of the causes that appear to be currently external to the competitive
arena (e.g., debt crises and globalized economies) are actually the effects of
prior policy decisions.
The claim that cartels will not form is directly related to Kitschelt’s third
complaint, that cartels are vulnerable to new entrants into the market (we
agree, see Chapter 7) and that it is not true that (p. 170) “party cartels
manage to prevent entry and, failing to do so, are able to coopt new parties
into the existing cartel, except those that make the new party cartels them-
selves the critical point of attack.” As noted above, the capacity of cartel
parties to prevent entry (or to handicap new entrants) is limited by the fact
that they are not all powerful. Likewise, the capacity of a cartel to coopt new
entrants depends on the willingness of the cooptee as well as the desires of
the coopter.
This points, however, to three more fundamental misunderstandings that
affect many of the criticisms of the cartel thesis. First, we never claimed that a
cartel of cartel parties would be stable; indeed, we argued exactly the opposite,
that the self-protective mechanisms of a party cartel would be unable
to prevent the emergence of challenges from outside the cartel . . . Thus in

much the same way as the elite parties created the social and political
conditions for the emergence and success of mass parties, and as the mass
parties, in turn, created the conditions for the emergence and success of
catch-all parties, and as the catch-all party led to the conditions that
generated the cartel party, so the more recent success of the cartel inevit-
ably generates its own opposition. (Katz and Mair 1995: 23–4)
Second, although we identified the cartel party with a particular time period
(Katz and Mair 1995: 18), we did not mean to imply that all parties in all
countries should be expected to be cartel parties in any full sense of the term.
Rather, for each of the models of party organization, we were suggesting that

there was/is a kind of “Zeitgeist” that would be especially conducive to one
model or another in a particular period, but that in each period other party
forms from earlier periods would continue to exist, and that new forms would
be in the process of emerging. Third, there is confusion of claims that parties
will pursue strategies with claims either that those strategies will be mutually
consistent or that they will be successful.
IMPLICATIONS
In the years subsequent to the publication of our original paper, the trends to
which we drew attention have become more easily seen, and serve to bolster
rather than weaken the overall argument. This is particularly so when we
look at the behavior of the established parties, which seems to come closer
and closer to the pattern we sketched, both in terms of party organizational
styles and patterns of competition. Moreover, regardless of whether one
accepts the cartel thesis in its entirety, it is evident that the growing incorp-
oration of parties within the state, their increasingly shared purpose and
identity, and the ever more visible gap that separates them from the wider
society, have contributed to provoking a degree of popular mistrust
and disaffection that is without precedent in the post-war experiences
of the long-established democracies. One may dispute the interpretation of
cartelization, but what is beyond dispute is the popularity of what is now
often identified as a populist, anti-cartel rhetoric. We will look at this issue
in Chapter 7.
One question that remains is where this leaves the concepts of party and of
party government—concepts that have been at the core of the understanding
of European democracy in particular and that we explore throughout this
volume. As suggested above, there are restraining factors that may limit the
degree to which parties follow the path we have identified. At the same time,
however, it seems unlikely that the parties would—or could—reverse their
drift towards the state, or that they could all somehow reinvigorate their
organizational presence on the ground.
It also seems unlikely that the parties—at least within the mainstream—will
discover some great issue divide or a new basis for policy polarization, and
when one remembers the bloodshed frequently associated with polarizing
questions of class or religion, it is not clear that it would be desirable if they
did. The neoliberal economic consensus is now well established in the minds
of mainstream political leaders, and on many of the issues that might offer the
basis for polarization in left-right terms the room for maneuver is either
limited, or the capacity to decide has been delegated elsewhere. This also
The Problem 23
seems to be the case even when parties have had to confront the worst effects
of the financial crisis after 2008. Beyond the economy and welfare, and
beyond the heavily constrained options available in fiscal and monetary
policy, there lie other issue dimensions that might serve to organize opposition
and that cut across the traditional class-based left-right divide. The environ-
ment offers one set of issues; immigration offers another; the international
order offers a third. But whether meaningful choices might be meaningfully
politicized in any of these issue areas, or whether, even if politicized, they
might offer the basis for widespread popular re-engagement in the electoral
process, is very much open to question. Moreover, even if such issues were
politicized and proved capable of stimulating popular re-engagement with
electoral politics, it is virtually unthinkable in modern societies that they
would be rooted in the kind of social cleavages that were a necessary condition
for the mass party model. For example, although Kriesi and his colleagues
(2008) are very emphatic in claiming to identify a new cleavage in European
politics shaped by the division between the winners and losers of globaliza-
tion, it is not at all clear that this conflict has found a consistent party political
expression, except perhaps in the support for new populist parties, or that it
can endure in the form of a stable alignment.
Much of contemporary debate concerning, and criticism of, parties and
party government, and much of the advice for building strong democracies in
the “third-wave” countries, and for addressing the “crisis of democracy” in
first- and second-wave countries, remains strongly informed by the mass party
model of ideologically/programmatically distinctive parties, each supported
by strong roots in society and governed internally by bottom-up democratic
practices. But at the same time, it is undeniable that for all practical purposes
the mass party is dead.
For now, it seems, we remain with a reality that is defined by a set of
mainstream parties that many perceive to be largely indistinguishable from
one another in terms of their main policy proposals, and that are closer to one
another in terms of their styles, location, and organizational culture than any
one of them is to the voters in the wider society. Elsewhere (Mair 2009), this
new configuration of party politics has been discussed in terms of the erosion
of the parties’ representative roles and the retention of their procedural roles,
and it has also been argued that in the absence of a capacity to combine both
roles, parties risk losing their legitimacy. That is, unless parties can represent
as well as govern, it may turn out to be more and more difficult for them to
legitimize their command of governmental institutions and appropriation of
public resources.
More immediately, however, these developments also raise the issue of
future models of party organization. To adopt Katz’s (1986) terms, the current
situation is characterized by an enhancement of the partyness of government—
as reflected in enhanced levels of party recruitment, nominations, and office

holding—but by a reduction of the partyness of society and party
governmentness—that is, reduction in the degree to which parties penetrate
the broader society and in the degree to which party government characterizes
the overall regulation of society. Within the institutions of government, party
organizations often dominate; within the wider society, the party presence has
been transformed into a professional electoral campaigning machine. The party
as campaigner attempts to reach out to as wide a range of voters as possible, but
the links that it establishes to these voters are at best contingent, instrumental,
and short term. They are also very direct, in the sense that the waning of the
party on the ground has left little or nothing in-between the competing sets of
leaders, on the one hand, and the available and often indifferent body of voters,
on the other. In this version of post-party democracy, there is little or no
mediation, and hence little or no role for traditional party organizations.
What lies between the elector and the elected is all but disappearing, rejected
by the disengaged voters, on the one hand, and by campaigning politicians, on
the other.
Whether this constitutes cause for alarm, or merely identifies a need to
adjust expectations and criteria for evaluation, is a matter of subjective
judgment. As Whiteley and his associates (2001: 786) observed—before
using the term themselves to describe the dramatic drop in voter turnout in
the 2000 British General Election—“[t]he word crisis is often abused in
contemporary accounts of politics.” Is a “crisis of democracy” that has
been going on for more than forty years (to date it from 1975, the publica-
tion of the Crozier et al. book of that title) really a crisis? At the same time,
popular disenchantment with democratic governments and with political
parties as the central actors in those governments, if not necessarily with the
abstract idea of “democracy,” is undeniably growing. To some extent,
disenchantment now is the result of unrealistic expectations (in part created
by the parties themselves) in the past, but whether or not prior expectations
were realistic, their disappointment has real consequences for the future. The
possible consequence on which most attention has focused is the rise of
populist, anti-party-system parties, and the danger that they will undermine
the liberal rights on which modern democracy rests. There is certainly
adequate precedent in the red scares of the 1950s, and the rise of fascist
regimes in the interwar period, to make this threat credible. But there is
another, and perhaps more insidious possible outcome. Rather than being
replaced by some kind of authoritarian, or perhaps initially authoritarian-lite,
regime, democracy might instead be hollowed out, becoming what Walter
Bagehot might have described as a “dignified part of the constitution”—a
revered legitimizing myth, but with little practical consequence in the actual
governance of society. But if the sphere in which the parties, including
the populist parties work, is marginalized, what power will move in to fill
the resulting void?
The Problem 25
PLAN OF THE BOOK
These developments then raise three important questions, or research agenda,

for scholars of political parties, and we seek to address—if not always to
answer—these questions in the present volume. Our first agenda, and the
major focus of this book, is to address more directly and fully the range of
empirical questions that have been raised with regard to the cartel thesis itself.
In Chapter 2, we trace the evolution of parties and party systems from the
mid-nineteenth century development of the mass party through the era of the
catch-all party in the last third of the twentieth century. Although the devel-
opments that together have led to the “crisis of party democracy” may be of
fairly recent origin, both the developments themselves and the set of expect-
ations that led to the diagnosis of crisis have much deeper roots.
Chapter 2 has two main objectives. The first is to emphasize the evolution-
ary nature of party change and that, in particular, the evolution is driven by
the rational adaptation of parties to social change, but also to the adaptive
behavior of other parties. As Marx (1852) observed, “Men make their own
history, but they do not make it as they please; they do not make it under self-
selected circumstances, but under circumstances existing already, given and
transmitted from the past.” If we are to understand party and party system
change going forward, we must first examine the processes through which
such changes have occurred in the past.
The second objective is to highlight the interconnection between party and
party system organization, on the one hand, and understandings of democ-
racy (including understandings of the nature of the demos), on the other.
Although it is tempting, and not entirely without truth, to assert that the
democratic theories articulated by political parties are little more than ration-
alizations for their own self-interested practices, once a democratic theory has
been claimed by a party to justify itself, that theory becomes a constraint on
future adaptation. In particular, because the appeal of the theory often
outlives the circumstances that made the party behavior that the theory
justifies attractive in the first place, the result can be a serious disjuncture
between expectations and practice that leads to widespread dissatisfaction.
One of the recurring themes of this book is that this is precisely what has
happened with the democratic theory of the mass party; although the mass
party form may have passed into history, the democratic theory that devel-
oped to justify it continues to influence both political science and quotidian
political discussion, to the detriment of current parties.
Although they are often treated as unitary actors, each political party is
actually a small political system in its own right. In Chapter 3 we develop this
point further, elaborating on the relations among “the three faces of party
organization”: the party on the ground, the party central office, and the party

in public office. The main themes in this chapter are the increasing dominance
of the party in public office and the increasing professionalization of the party
in public office and the party in central office—both the development of a
separate vocation of politics and the replacement of people who “live for
politics” with people who “live off politics” (Weber 1919).
Chapters 4 and 5 are concerned with relations among parties (Chapter 4)
and between parties and the state (Chapter 5). In both cases, this is a story
about moving together. On the one hand, the parties have tended to become
more similar to one another. There is less differentiation among their policy
proposals in elections and the policies they pursue in office; their methodolo-
gies in electoral competition have become more similar, as have their organ-
izational structures. On the other hand, the distinction between the parties as
tools of civil society attempting to secure temporary control over the state,
and the state itself, has become increasingly difficult to maintain; rather than
being private actors attempting to control the state apparatus, the parties
increasingly have become part of it.
The joint result of these processes has been the gradual and partial evolu-
tion of a new model of party organization, a new pattern of “normal”
relations among parties, and a new set of legitimizing principles, that come
together to define what we have identified as the “cartel party.” In Chapter 6
we both develop these ideas and, building on the evidence presented in earlier
chapters, advance the claim that parties are in fact evolving organizationally
and behaving in ways that reasonably approximate the cartel ideal type.9
Given the two-pronged nature of the cartel party argument—that is, a cartel-
ized party system and individual cartel parties within that system—this
requires us to confront two sets of questions. On the one hand, we develop
and assess indicators of the cartelization of the party system, for example,
addressing whether we see a constriction of competition among cartel mem-
bers and increasing rules that advantage cartel members over those outside
the cartel. Whether or not there is evidence of actual collusion among the
parties, do we find evidence of the behavior that we would expect if there were
collusion? On the other hand, we look inside parties for evidence that their
organizations and practices are coming into line with the predictions of the
cartel model: more dependence on state resources; greater emphasis on
improved management rather than reformed policy; more formal but less
substantive internal democracy.
Finally, in Chapter 7 we address the question of what difference this
all makes for the future of democracy. As we have already observed, there
9
Again, approximation to an ideal type is all that can be claimed for the real-world mass
parties or catch-all parties. Indeed, because each step in the evolution of party types has
stimulated the development of a countervailing form, failure of real parties fully to conform to
any of these ideal types is actually part of the model.
The Problem 27
is—and has been for at least forty years—widespread talk of a “crisis of
[party] democracy.” Two of the contemporary manifestations of this crisis
are the increasing withdrawal of citizens from involvement with the main-
stream parties, and the concurrent rise in support for radical populist parties,
generally but not exclusively of the right. The cartelization of mainstream
party politics is clearly implicated in these processes, both as cause (main-
stream parties failing adequately to represent the perceived interests of citi-
zens) and as effect (all the mainstream parties “circling the wagons” and
turning to the state for support in the face of declining support from their
erstwhile base). The overall result is a growing disjuncture between popular
expectations regarding parties and their actual performance.
One clear danger, which fortunately does not yet appear to have material-
ized, is disenchantment with democracy tout court. While it may be exces-
sively alarmist to see the populists as harbingers of a return to fascism, the
possibility that liberal democracy will be supplanted by some fundamentalist
(whether religious or not) ideology that promises to protect the interests of the
people against the corrupt and corrupting elite cannot be entirely discounted.
If the gap between performance and expectations continues to grow, the
danger of reaching the breaking point will grow as well.
One strategy suggested for closing the gap between performance and
expectations lies in the emphasis in the “New Public Management” school
for improved “customer service,” taking the supposed “customer responsive-
ness” of the private sector as its point of reference (Osborne and Gaebler 1992:
Barzelay 1992). In this scenario, citizens as active participants in their own
government are transformed into consumers of government services. While
initially this idea was advanced as a prescription, more recently it has also
been suggested as a description of what governments actually are doing—
whether by intent or as an unintended, but nonetheless real, consequence (e.g.,
Mosse and Whitley 2009). But as many critics have pointed out, the relation-
ship of citizen to state is not the same as the relationship of customer to firm.
The state is a monopoly supplier with the power of compulsion, in both
respects denying to the citizen the option of exit that is characteristic of
most private-sector transactions. The relationship of consumer to firm is
individual and concerned with private goods, while that of citizen to the
state is often collective and concerned with public goods (whether policies
or material goods). The private sector is characterized by a direct connection
between delivery of services and payment for those services; the public sector
is not (Pegnato 1997). Thus, even if the goals of the New Public Management
were achieved, this would likely only reduce the gap between expectations and
performance with regard to individual interactions with the state and the
delivery of personal services. It would be far less likely to ameliorate dissat-
isfaction concerning the content of policy, the constriction of the range of
options offered to voters, or the general quality of democracy.

Although it is possible that the gap between performance and expectations
will be reduced by improved performance, the trends that we document in this
book are all in the opposite direction. Eurobarometer data show a substantial
(11 percent), and nearly monotonic, drop in the proportion of respondents
“satisfied with the way democracy works” in their country between Autumn
2004 and Autumn 2013, albeit with a 4 percent recovery in Spring 2014 (EB81
Report). There is, however, an alternative way in which the gap can be
reduced: that is, by lowering expectations of what governments can achieve
or be held accountable for. A not-always-successful attempt to do this is
already evident in the offloading of responsibilities to independent central
banks, the EU, or the WTO. Although generally proposed as ways of improv-
ing performance, the parties also can hope that they will lower expectations
for the parties themselves—because they are at worst indirectly responsible
for the outcomes, and at best not held responsible at all. In this case, we have
an evolution of liberal democracy in which the liberalism is heightened and
the democracy is hollowed out. Rather than government by the people, it
becomes government for the people by civil servants and technocrats. The role
of elections is reduced (in the words of British satirists Antony Jay and
Jonathan Lynn (1988)) to “deciding which bunch of buffoons will try to
interfere with our [the technocrats’] policies.”
Notwithstanding obvious differences, the result would bear many resem-
blances to the practices of earlier centuries in which the monarch (now
replaced by bureaucrats and associated experts) ruled, and the people’s rep-
resentatives voiced grievances but did not exercise substantial power. As
Schattschneider observed in the passage quoted in the introduction to this
chapter, political parties played a crucial role in the transition from monarchic
government to democratic parliamentary government. We now turn to the
question of how parties evolved after they became the central players in
liberal, and then in liberal democratic, government.
The Rise and Decline of Parties
In this book we address the contention that many Western democracies are
now experiencing a major crisis of party government and party democracy.
How exactly one defines the concepts of party government or party democ-
racy, and whether they denote the best or only possible form of democratic
government, or whether alternatively they denote only a pale shadow of real
democracy, are, of course, subject to debate. For some, little more is required
for democracy than that the principal offices of state be filled by contested
elections meeting quite minimal standards of fairness (e.g., Collier and Le-
vitsky 1997: 440), while for others this is at best “thin” (Barber 1984) or
“elitist” (Bachrach 1967) democracy. But at the core of both party govern-
ment and party democracy is the notion that political representation and
authority within democracies are and should be channeled through the
medium of party. In a system of party government, adopting the criteria
specified by Katz (1987), political decisions are made by elected party officials
or by those under their control; policy is decided within parties, which then act
decisively to enact these policies; and, finally, public officials are recruited
through and held accountable by parties, or else are controlled by those who
are so recruited and accountable. Party democracy is less tightly defined, and
in the literature is often taken to refer to democracy within parties rather than
to the role of parties within democracy at the system level. Indeed, in this
latter and wider sense, it is rarely even discussed. Bernard Manin (1997), who
adopts the term as a cross between the English “party government” and the
German “Parteiendemokratie” (p. 197, fn. 6), speaks of it as a system in which
“people vote for a party rather than for a person” (p. 208), and in which
“parties organize both the electoral competition and the expression of public
opinion (demonstrations, petitions, press campaigns)” (p. 215), thereby laying
a welcome emphasis on the role of parties within the wider democratic
process.1 Similarly, in contrasting “party democracy” to both populism and
technocracy, Bickerton and Invernizzi Accetti (2017) define it as
1
Given that in one of the archetypical cases of party government, the UK, people technically
vote for a person rather than a party (until 1998 officially recognized party names did not even
appear on the parliamentary election ballot), the first of Manin’s conditions clearly has to be
understood subjectively.

a regime based on two key features: the mediation of political conflicts
through the institution of political parties; and the idea that the specific
conception of the common good that ought to prevail and therefore be
translated into public policy is the one that is constructed through the demo-
cratic procedures of parliamentary deliberation and electoral competition.
Since party is central to our current understanding and learned practice of

democracy, a crisis of party government and party democracy can be seen to
have major implications for democracy itself. At the very least, a questioning
of the legitimacy and capacity of parties to represent citizens, on the one hand,
and to manage the institutions of government, on the other—the two key
functions of the democratic polity that were fused through the medium of
political parties—obliges us to rethink many of the models and assumptions
regarding the good functioning of twenty-first-century democracy. Without
parties, and indeed without legitimate parties, how would democracy and
representative government be sustained? Without parties, would citizens still
be able to find a voice within the polity and able to hold their governments
accountable? Is partyless democracy, to paraphrase Schattschneider (1942: 1),
genuinely unthinkable?
Given that talk about a “crisis of party” or a “crisis of democracy” has been
common for half a century, our concerns may seem excessive. More than forty
years ago, democracy was also seen to be in crisis, but at that point it was
believed to be the result of a combination of government overload, increased
social mobilization, and a popular demand for greater equality and participa-
tion (Crozier et al 1975). Since then, as Pharr and Putnam (2000: 3–27) remind
us in their introduction to a more recent re-evaluation of these questions, we
have witnessed the end of the Cold War, the global resurgence of democracy,
and—at least then—a period of unprecedented material prosperity. Rather
than excessive demands burdening governors, it began to seem that citizens
were becoming indifferent to politics and were increasingly likely to withdraw
into their own well-furnished private spheres. As Galbraith (1992) had sug-
gested in the American case, the earlier crisis of democracy appeared to have
been succeeded by “a culture of contentment.” The idea that this has now been
followed by a new crisis of democracy might therefore seem equally misplaced.
Perhaps either the problems will be ameliorated, or the parties’ capacity to
cope with them will increase, so that the current crisis is supplanted by another
period of relative satisfaction only to be followed by a new crisis, in an
unending stream of “crises,” none of which ultimately proves to be critical.
We do not take much comfort in this possibility, for reasons that we will
develop in this and the chapters that follow. One reason why one might think
that our projection is too dire, however, would be the claim that in pointing to
problems faced by party democracy and party government today, we under-
estimate the extent to which these or other problems also confronted parties in
the earlier periods. As Scarrow (in Dalton and Wattenberg 2000; and in more
detail in Scarrow 2015) has suggested, arguments such as ours risk assuming
The Rise and Decline of Parties 31

that there was some golden age of party in the past, or rely on the notion
that we have exited from some mythical period in which party democracy,
party government, and representative politics functioned smoothly within
an environment of wholesale popular legitimacy, widespread popular trust,
and robust and well-populated party organizations. Such assumptions
would, of course, be misleading. Parties were also troubled in the past,
and cracks could sometimes be found in the frame of party government.
Parties and their leaders frequently were accused of petty venality, if not
wholesale corruption. At least some parties had problems with inactive
membership and with poorly developed links with civil society. Indeed,
some parties simply failed, or came close to failing. But while this some-
times appears true in the specifics, it is not true in general. As we argue in
this chapter, and as we seek to show by tracing the development of party
organizations over time, the past was better for parties in terms of organ-
izational strength and legitimacy, as well as in terms of their capacity to
manage party government and to sustain party democracy. The past may
not have been a golden age, and it may not have proved uniformly better for
all of the parties that then competed, but it is nevertheless clear that the
position of parties in the early post-war decades was generally more secure,
more stable, and more effectively embedded in democracy than is now
the case.
The classic party government model portrayed in Figure 1.1 (and elabor-
ated in Figures 1.2 and 1.3) was always an ideal type rather than a precise
description. Nevertheless, it is our contention that a confluence of social
changes, institutional and environmental changes, and changes in the parties
themselves has moved reality farther away from the ideal type, ultimately to
such an extent that, as we argue later, it becomes useful to propose and outline
a new model of party. There are few if any dramatic conjunctures in this
evolutionary trend—although one that did matter was the short three-year
span between 1989, when the Cold War effectively ended, and 1992, when the
European single market was established—but instead a relatively slow process
of gradual adaptation to change and of attempts by parties to control their
environments, a process that still continues today. But in order to understand
this process, and its contemporary products, it is necessary to look more
closely at how the models of party evolved in the past decades, and how
that dynamic might be explained.
THE ELITE PARTY
Although, as with nearly all things, the roots of political party can be traced
into the mists of antiquity, the story of modern parties essentially begins in the

parliaments of Europe, and in a decidedly pre-democratic era of royal power.
Even royal absolutism had limits, and even more in countries where royal
power was constrained by tradition or countervailing sources of authority—or
where, as in the Republic of the Seven United Netherlands or in Switzerland,
there was no king—it became common quite early (as early as 1213 in
England, for example) for assemblies of representatives of the country to be
called, both to provide local information and to articulate grievances, on the
one hand, and to grant or consent to taxes or even to legislate, on the other.
These early parliaments in what would become the liberal states of Northern
Europe were composed of representatives of local communities. Even when
representatives were nominally elected, although in many cases “appointed”
would be a more accurate description, suffrage was highly restricted, and
elections were commonly “arranged” by the local elite. For example, in the
election of 1761 in Great Britain only 48 out of 315 constituencies actually
went to the polls, a number of contests which Namier reports to have been “no
more numerous than usual” (Namier 1963: 159), and well under 10 percent of
adult males could vote. Especially in these cases, and even in those cases in
which there was a contest, campaigns and the associated organizations
remained fundamentally local and independent exercises in the mobilization
of locally based support. Even into the nineteenth century, there was little
need, or desire, for centralized coordination or control beyond such things as
the scheduling of speaking tours by national figures.
Under these circumstances, political organization in the constituencies
(“the party on the ground”) could be little more than the social network of
retainers and clients of the member of parliament—or in other cases, the
other clients of “the duke or lord or baronet whose representative he [was]”
(Beer 1969: 23). By the mid-nineteenth century, suffrage expansion meant
that there were some constituencies in which more extensive organization
was required in order to mobilize a larger electorate. For example, after the
Reform Act of 1867, the electorate of the British constituency of Birming-
ham increased from 7,309 in 1832 (almost 15,000 in 1865) to over 42,000 in
1868 and over 55,000 in 1874 (Vincent and Stenton 1971: 23), providing a
strong incentive for the evolution of the “Birmingham caucus” as an early
constituency-based party organization.2 Even before the suffrage expan-
sions of the nineteenth and early twentieth centuries, however, some of the
conditions that had been conducive to the elite parties began to change. The
expansion of the role of government (Burke 1770) and the development of
2
Exceptionally, France and Switzerland have had universal manhood suffrage since 1848, as
has Germany since 1867 (in the North German Federation) or 1871 (in the German Empire).
Women were enfranchised on the same terms as men in Germany in 1919, but not in France until
1944 and in Switzerland until 1971 (for federal elections, and 1990 in the canton of Appenzell
Innerrhoden).

notions of government responsibility to parliament (Jennings 1969: 17–18;
Sartori 1976: 18–24) increased the value of reliable party cohesion within the
party in public office. It also increased the national relevance of local
elections, stimulating greater communication and coordination across the
local parties on the ground (see Caramani 2004, especially for developments
from the mid-nineteenth century).
As parliaments evolved from assemblies of local ambassadors effectively
making deals with their sovereign, exchanging consent to taxes for redress of
grievances,3 into collective makers of laws and supporters (and then choosers)
of ministries, the ability to muster a reliable majority within parliament
became crucial to ministerial success. Moreover, even those who did not
aspire to ministerial office learned that by coordinating their voices and
votes, a group of political “friends” could exercise far more influence than
the same representatives acting individually. By supporting a single spokes-
man, and giving him the capacity to speak for and commit the entire group,
over the long haul even those who were not entirely in agreement with their
fellows could expect outcomes closer to their own preferences than would
result from independent action. Over time these spokesmen became more
clearly recognized as leaders, with the capacity to demand and reward loyalty,
and to punish defection—and the groups of political friends became clearly
identifiable as parliamentary parties.
There was another crucial development associated with the rise of parlia-
mentary groups: the acceptance of the idea that parliamentary opposition was
legitimate rather than traitorous. At least through the sixteenth century,
parliamentary groups would generally have been identified as “factions”—
that is, as Madison (1961 [1787]) put it in The Federalist no. 10, as “a number
of citizens, whether amounting to a minority or majority of the whole, who
are united and actuated by some common impulse of passion, or of interest,
adverse to the rights of other citizens, or to the permanent and aggregate
interests of the community.” By the middle of the nineteenth century,
Edmund Burke’s (1770) definition of party as “a body of men united for
promoting by their joint endeavours the national interest upon some parti-
cular principle in which they are all agreed” held sway, at least in England.4
This transformation was, no doubt, facilitated by the fact that no matter
how intense the disagreements among members of parliaments may have
3
Notwithstanding Edmund Burke’s eighteenth-century claim that the British parliament was
“not a congress of ambassadors,” early English writs of election called for MPs to be sent to meet
with the same plena potestas that a sovereign might grant to his ambassadors.
4
In Germany, for example, Heinrich von Treitschke argued against the British model of party
government, seeing parties (as would become the dominant theme in the work of twentieth-
century scholars like Joseph Schumpeter and Anthony Downs) as dominated by a drive to rule
rather than shared ideas. See the brief selection in Scarrow (2002: 159).

appeared, they were all drawn from a relatively small set of social strata, who
shared a common interest in maintaining their own ascendency against chal-
lenges from those much larger, and increasingly vocal, classes that were
excluded from legitimate participation. In many respects, the difference
between legitimate dissent and sedition lies in who is doing the dissenting.
The régimes censitaires in which these elite parties developed were hardly
democratic, unless one defines the demos to include only those who qualified
for the vote. Viewed from within that restricted demos, there appeared to be
real disagreements and real competition—although sufficiently constrained
that it was plausible for Lady Bracknell (in Oscar Wilde’s The Importance of
Being Earnest) to say of Liberal Unionists that “they count as Tories. They
dine with us. Or come in the evening, at any rate.” As the political conscious-
ness of those outside of that circle grew, however, it would have been perfectly
plausible for them to see the system, and all of the elite parties operating
within it, not so much in competition with one another as cooperating
(colluding) with one another to maintain the social and political status quo,
and their own privileged position within it.
Obviously, the normative theory undergirding this pre-democratic system
was more liberal than democratic, but several elements of it remain relevant
even after both the elite party and the social system in which it existed have
disappeared. First, it was a theory founded on the idea of a single national
interest and that the objective of politics was to advance that interest. While
there might be disagreement about its specific content and about the best
means of discovering or advancing it, there was a common interest that was
normatively superior to any private interests. Second, while recognizing the
importance of coordinated action, it believed in the importance of the
independent judgment of individual office holders and in the importance
of rational deliberation.5 Thus while it valued party cohesion, it opposed
party discipline. Third, while not always justifying a highly restrictive
suffrage (in the British case, for example, both Jeremy Bentham and
John Stuart Mill advocated nearly universal manhood suffrage, and the
British franchise was greatly expanded under Whig (1832), Tory (1867),
and Liberal (1884) governments), it remained elitist both in assuming that
while “the people” might be competent to choose who would govern, they
were not competent to govern themselves, and in assuming a high level of
deference to, and public-spirited responsibility from, those who formed the
political elite.
5
Although it should not be taken as the definitive articulation of the political theory of the age,
this is well summarized in Edmund Burke’s famous address to the electors of Bristol, November 3,
1774.

THE MASS PARTY
While the elite party fit well with the traditional rural societies and limited
suffrage of the pre-industrial era, with the rise of industry and especially of
commerce organized on a national scale, improved communication, and
growing urbanization, by the later decades of the nineteenth century this
was no longer the case. Before the 1840s, for example, it would take almost
two weeks to travel from Edinburgh to London by stage coach; after the
railroad connection was completed in 1848, travel time between Edinburgh
and London was reduced to 12.5 hours by train—and of course even if
individual people rarely made the journey, newspapers and mail did so on a
daily basis. Similarly, the somewhat longer (in distance) journey from Paris to
Marseille in 1789 would take nine days by stage coach (after major road
improvements in the eighteenth century (Roche 1998: 55)), or more than three
days in 1830, but only sixteen hours after the railway was completed in 1856,6
and closer to eight hours with the introduction of Crampton locomotives after
1864. Coupled with the invention of the telegraph in the 1830s and the
telephone in the 1870s, close coordination between politicians in the capital
and their supporters or constituents in the rest of the country became much
more possible.
Along with changes in communication, the nineteenth century also wit-
nessed substantial social change. By one estimate, in Western Europe in 1800,
only Belgium (20.5 percent) and the Netherlands (37.4 percent) had more than
20 percent of their populations living in cities of 5,000 or more inhabitants; by
1900 only Finland, Portugal, and Sweden were under 20 percent urban—and
the UK was more than two thirds urban (Bairoch and Goertz (1986); specif-
ically on England and Wales, see Law (1967: 130)). Along with urbanization
came the growth of trade and industry, and the rise both of a commercial elite
and of a large non-agricultural working class.
While there clearly were intense conflicts between the old aristocracy of
land and the rising commercial aristocracy, illustrated in Britain, for example,
by the struggle over the repeal of the corn laws, the rising elite not infrequently
intermarried with the traditional aristocracy and politically could be inte-
grated into the existing elite parties. While those parties became more nation-
alized and more centralized, and in some cases developed a more than
rudimentary extraparliamentary presence, they were still primarily located
in parliament and while still claiming to be promoting the national interest in
6
Similarly, in 1789, it took five days from Paris to Lyons by stagecoach (up to fifteen days to
send a package (Roche 1998: 55), but according to the Commission d’enquête sur l’exploitation et
la construction des chemins de fer (1858: xiv), the average express train made the run in eight hours
and twenty-four minutes in the 1850s.

fact represented what was numerically only a fraction of the population.
Whether “notability” was based on commercial fortune or position in
the landed gentry, these remained parties of notables whose independent
positions did not rest primarily on the holding of public office.
Integration of the rising working class was an entirely different matter. The
French Revolution of 1789, and then the unrest of 1832 and 1848, appeared to
pose a common threat to both landed and commercial elites. The incorpor-
ation of the commercial class was accomplished under régimes censitaires,
which excluded the working class, and whether and how much to lower the
barriers to electoral participation became a major issue both among the
parties representing those already enfranchised, and between them and emer-
ging organizations of the unenfranchised—trade unions in particular. Indeed,
enfranchisement was often one of the objectives of the trade unions, along
with better working conditions and higher wages.
Whether greatly expanded suffrage was achieved rapidly or gradually, by
the end of the First World War, (nearly) universal manhood suffrage was the
norm in Europe. With the expansion of the electorate from thousands to
hundreds of thousands, raw numbers became a valuable political resource,
and at the same time more elaborate organization became a necessity. For
those interests whose potential strength lay in numbers of supporters rather
than in the “quality” of their individual supporters, notably the working class
and, in some cases (such as the Netherlands) fundamentalist Protestants, the
elite party model clearly was inappropriate. Archetypically, the parties that
developed to represent and advance these groups initially had no party in
public office, because the majority of those they aspired to represent were
excluded from electoral participation. Even if their core organizers included a
few members of parliament elected through one of the older parties, they
perceived one primary task to be the formation of independent organizations
that would mobilize their supporters, first to win the right to vote, and then to
provide both the votes and the other resources required to win elections under
the new conditions of mass suffrage.7
Because these resources had to be amassed on the basis of many small
contributions from ordinary people rather than coming from a few wealthy or
powerful individuals, this effort required a substantial, and well-organized,
party on the ground. And because the demands of these groups involved
fundamental changes in national policy, it also required organization and
coordination across constituencies, that is to say a substantial organizational
capacity at the national level, and hence a strong central office. Both of these
7
Although it is only marginally relevant to the main thrust of our argument, it is worth noting
that there was considerable temporal variation in Europe with regard to the extension of freedom
of association, the legalization of trade unions, and the development of membership-based parties
(Scarrow 2015: 44, 56).

requirements were heightened in the case of many of the more ambitious and
“isolated” parties by a strategy of encapsulation, which required the main-
tenance of a panoply of ancillary organizations, and by the fusion of electoral
mobilization with additional activities. The organizational form that evolved
to meet these needs is the mass party (Duverger 1959 [1951], 63–71), and
within the mass party it is the need to manage the organization and to
coordinate its activities that leads to the development and consolidation of
the party in central office.
Whether the party in central office was formed first for the purpose of
creating a party on the ground, or was formed as an umbrella for the
political/electoral activities of previously existing organizations (e.g., churches
or trade unions), is less significant than the symbiotic relationship between the
two. The party in central office provides support for the expansion of the party
on the ground and central coordination for its activities, while the party on the
ground provides the resources that are necessary for the existence and success
of the party in central office. As in any symbiotic relationship, it is difficult to
say whether the party in central office or the party on the ground will be
dominant, or even what dominance would mean.
In the ideology and formal structure of the mass party, the party in central
office is the agent of the party on the ground (Beer 1969, ch. 3). Its leading
officials are elected at a party congress as the representatives of the mass
membership. But having been elected by the members, and therefore occupying
a position presumably subservient to the party on the ground, the leaders of the
party in central office also have been given a mandate to manage the party, and
presumably to make rules for and give directives to the party on the ground
(McKenzie 1963). It is particularly in this nexus that questions about internal
party democracy and the iron law of oligarchy (Michels 1962 [1911]) are raised.
While the power relationship between the party in central office and the
party on the ground is somewhat ambiguous, the fact that these two faces are
separate is perfectly clear. The party in central office is staffed by full-time
professionals and administrators; the party on the ground is overwhelmingly
made up of part-time volunteers. People in the party in central office are paid
to be members; people in the party on the ground generally must pay in order
to be members. The party in central office and the party on the ground are
likely to be motivated by different varieties of incentives, and to measure
success by different standards (Panebianco 1988: 9–11, 24–5, and 30–2).
Nonetheless, their relationship can be fundamentally harmonious. Even
where the party in central office is clearly dominant, it claims to exercise
this dominance in the name of the party on the ground, while to the degree
that the party becomes a single national entity, dominance by the party on the
ground can be exercised only through a strong party in central office.
The mass-party model also clearly separates the party on the ground from
the party in public office. No longer an informal caucus of a few individuals,

the party on the ground grows to include hundreds, if not thousands, of
members. The member of parliament no longer can be seen as simply one of
the party elite taking/serving his turn, but assumes a distinct organizational
role. Moreover, within the ideology of the mass party, the role of the member
of parliament, and hence that of the party in public office, is clearly to be
subordinate to the membership organization. In the elite party, party organ-
ization is instrumental to the achievement of the goals of the individual
members of the party in public office. In the mass party, the party in public
office is instrumental to the achievement of the goals of the party organiza-
tion. In this respect, the party in central office has another crucial function,
that of supervising and controlling the party in public office on behalf of the
party on the ground.
The mass-party model is the first in which there is a clearly observable
empirical distinction among the three faces of party organization (distinct and
separate organizational presences; made up of different types of people;
different and potentially conflicting incentive structures) rather than the
distinction existing only at the theoretical/conceptual level. It implies a par-
ticular organizational form (local membership branches supplemented by
ancillary organizations; a representative party congress electing a central
party executive; etc.), but it also depends on a particular balance among the
three faces. In the early days of the mass party, and generally in the early days
of any party organized in this fashion, the party in central office, whether
acting independently or as the real agent of the party on the ground, was likely
to be dominant. It controlled the resources, and the party in public office, its
potential rival for dominance, had not yet experienced either the demands or
the rewards of control over the government. Once the latter did gain access to
the resources of government, however, it was always likely to assert greater
independence, and thus threaten the “mass partyness” of the organization. In
this sense, the mass party as an organizational type is naturally a party of
opposition; even if an individual mass party might win the occasional election
and temporarily be in power, long-term control over the government would
seriously undermine the internal balance among the three faces that is essen-
tial to the mass-party model. This is something we shall explore at greater
depth in Chapter 3, but here we can recall the finding of Robert McKenzie’s
(1963) classic study of power within British political parties, that while the
Labour Party adhered reasonably well to the mass-party model when it was in
opposition, this was not the case when it was in government.
In its extreme form, as both Gunther Roth (1963) and Sigmund Neumann
(1956) emphasized, the mass party, whether based on class or denomination,
became a world in itself, isolated from the rest of the polity, and threatening
rather than competing with its opponents. In its less extreme version, princi-
pally reflected in the form of large social democratic and Christian democratic
parties, it emphasized the engagement and activation of the members, and

played a key part in, perhaps even controlling, the network of ancillary
organizations that constitute the class or religious cleavage base. The key
motifs in such a context are solidarity, self-defense, and representation, and
it was in this sense that party loyalties could grow and that party
identification—or cleavage identification—could strengthen.8 Though not a
wholly closed political world in this more moderate version, it was neverthe-
less one in which many voting preferences could be taken as given, and in
which, as Rose and Mossawir (1967: 186) once put it, “to speak of the
majority of voters at a given election as ‘choosing’ a party [would be] nearly
as misleading as speaking of a worshipper on a Sunday ‘choosing’ to go to an
Anglican, rather than a Presbyterian or Baptist church.”
As Lipset and Rokkan (1967) argue, the cleavage structures in which mass
parties arose and embedded themselves did not just happen; they were the
consequence of deliberate strategic choices by elites engaged in, or opposing,
the process of modern state formation. This happened before mass suffrage,
and therefore generally before the advent of mass parties (although on the
cusp of mass enfranchisement organizations that would become mass parties
may have agitated for suffrage expansion). Once formed, however, mass
parties had a strong incentive to deepen cleavages that worked to their
advantage (e.g., the class cleavage for socialist and social democratic parties)
by encapsulating their supporters in an all-encompassing network of subcul-
tural organizations. They (and their non-mass-party competitors) also had an
incentive to try to break down cleavages that did not work to their advantage.
Thus Christian Democratic parties might emphasize the religious cleavage,
and create their own organizations for workers—for example, the Associa-
zioni Cristiane dei Lavoratori Italiani—in which Christianity rather than
social class would be the focus. Similarly, the British Conservative
Working-Men’s Clubs (Ball 1995: 8) provided an alternative to the “chapel”
and then to the socialist subcultures that supported the Tories’ opponents.
Houska’s (1985: 33–56) analysis of political subcultures in the Netherlands
and Austria in the 1950s and 1960s is a telling reminder of the importance of
such cleavage-based and segmented party worlds. To be sure, these were not
just party worlds but also sometimes religious worlds, and hence there already
existed a tradition of segmentation and separation (e.g., Kalyvas 1996:
63–76). But even when this was the case, parties came to be central to the
process. In the Netherlands, for example, almost 100 percent of the trade
union members who supported the Catholic party were members of Catholic
8
One explanation for the relatively low levels of “party identification” (answers to questions
equivalent to the American “Generally speaking, do you think of yourself as a Democrat, a
Republican, an Independent, or what?”) found in early European surveys was that class or other
cleavage identification served the same function (Shively 1979), with party choice following
unproblematically from that.

trade unions, while 95 percent of party supporters also supported Catholic
broadcasting associations. Almost 100 percent had attended Catholic primary
schools, and 65 percent read a Catholic daily newspaper. Within the Protest-
ant party camps the corresponding numbers were almost identical, while
among Labour Party supporters the situation was only slightly more muted:
87 percent of Labour-supporting trade union members were members of
socialist trade unions, 64 percent read a socialist daily newspaper, and 73
percent had attended non-religious primary schools. In Austria, where party
membership levels have always been among the highest in Europe, the seg-
mentation was even more pronounced. Almost 1 million Austrians—some 15
percent of the population—were members of party-affiliated sports clubs, and
almost half that number were members of party-backed pensioners’ associ-
ations (Houska 1985: 46). In Austria, in other words, as in the Netherlands,
the mass parties played a central and sometimes defining role in what was
clearly a very dense cleavage structure.
Beyond these two cases, the phenomenon defined by Lorwin (1971: 142) as
“segmented pluralism,” and that he deemed characteristic of systems in which
“cleavages have produced competing networks of schools, communications
media, interest groups, leisure time associations, and political parties along
segmented lines, of both religious and antireligious nature,” also flourished in
Belgium, Luxembourg, and Switzerland, probably reaching its apogee in the
1950s (Lorwin 1971: 163). Indeed, what is striking about Lorwin’s depiction
of this system is that while he emphasizes the central role of mass parties in
what was a complex of cleavage organizations, he also makes clear that the
party was just one actor among many. The other actors included the churches,
unions, schools, and the many professional as well as recreational associ-
ations, all of which played a role in sustaining the cleavage as a distinct
subculture. Italy, where the apogee of segmentation was probably reached
in the 1970s, somewhat later than in the consociational polities, offered
another example of strong segmentation, with the Catholic and communist
subcultures remaining sharply divided from one another from the 1940s
through to the end of the 1980s, and with both being sustained, albeit in
steadily weakening fashion, by a myriad of other organizations and associ-
ations (e.g., Farneti 1985: 160–1, 175–8). Nor was this only a phenomenon in
those polities where religious parties played a prominent role. Rustow, for
example, pointed to similar processes in the more secular Swedish environ-
ment of the early 1950s, while also emphasizing the central position played by
the party in particular:
A party in Sweden as elsewhere in Europe is not only a political appar-

atus; it is also a civic club, a pressure group, and an organization for the
pursuit of various leisure-time interests. The parties have separate wo-
men’s and youth organizations; they sponsor boy scout groups, summer

camps, civic training centers for party workers and others interested in
public affairs, and adult education classes in subjects as non-political as
mathematics and foreign languages . . . A politically conscious citizen
therefore does not just “vote for” or “register in” a party at stated
intervals: he is a Socialist or Agrarian, a Liberal or Conservative. In the
national party organizations the welter of proliferating activities finds a
center of coordination, and through their leaders and parliamentary
groups the organizations become articulate in national politics.
(Rustow 1955: 144–5)
Divisions and patterns of segmentation in other European polities were not

always so pronounced, of course. In Norway in the 1950s, the parties tended
to keep themselves at one remove from the panoply of interest associations
and groups that operated at both local and national levels, and were already
losing members and activists by the end of that decade (Valen and Katz 1964;
Torgersen 1970). In France, despite the sporadic intensification of the
religious–secular divide, it has always been difficult to speak of anything
approaching a political subculture, and the parties themselves never seemed
willing or able to develop strong mass organizations (e.g., Bartolini 2000,
428–36). Although British parties never generated or became part of distinct
political subcultures in the manner of Lorwin’s segmented polities, they were
nevertheless sustained by powerful class identities (for a recent evaluation,
see McKibbin 2010), and in this sense the cleavage structure was strongly
rooted. Moreover, both parties emphasized the need for a large and often
active membership. In 1953, for example, the Conservatives claimed close to
3 million members among a total electorate of some 35 million, a remarkable
degree of organizational presence, while Labour, though smaller in terms of
direct membership, claimed an indirect membership of more than 5 million,
the vast majority of which was based on the interpenetration of the party and
the trade unions (Epstein 1967: 112–13);9 indeed, in 1960 twelve of the twenty-
eight members of Labour’s National Executive Committee were specifically
chosen as trade union representatives (Webb 1992: 855).
Whether this period represented a “golden age” of party is a matter of
debate and further investigation. What is clear, however, is that this was a
distinct period in party development in which many parties maintained large-
scale and highly extensive mass organizations that were often embedded
within a complex of associated and affiliated organizations and activities that
made up a more or less closed cleavage community. The pattern was not
9
As Scarrow (2015) points out, party membership figures must be approached with caution,
both because there was no national registration of members and because there often was an
internal party incentive for local branches to exaggerate, or even fabricate, membership figures.
For our purposes here, however, it is evidence of the importance the parties gave to high
membership numbers, rather than the accuracy of the specific figures, that is important.

necessarily evenly distributed across the entire party spectrum within any
polity, and this sort of mass party also competed with other parties that did
not enjoy similar structural foundations. This led to two important conse-
quences. On the one hand, the distinct and closed social bases of the mass
parties led to strong party identities and to distinct differences in policies, even
among competing mass parties. On the other hand, because not all parties
were mass parties, or had this penumbra of ancillary organizations, parties
might be very different in their organizational arrangements as well.
Mass parties, particularly when embedded within strong political subcul-
tures, helped to integrate citizens into political participation and into the
political system more widely, while at the same time they helped to isolate
one group of citizens from another and thereby helped reinforce and “freeze”
the cleavage structure. The most apt metaphor, as Rokkan (1977) also once
observed, is verzuiling or pillarization, and comes from the consociational
Netherlands: in such cleavage-based systems, each of the main parties consti-
tuted the political leadership of a vertically integrated pillar, or cleavage
group, and these pillars grew alongside one another without much mutual
interaction or cross-cutting pressures. At the top of the pillars, in the parlia-
mentary and governmental arenas, negotiation, accommodation, and even
formal coalition making across the pillars made peaceful governance possible,
and as the interwar period demonstrated, the absence of such negotiation and
accommodation at the peak of the pillars made democratic collapse likely.
Within each pillar itself, however, isolation proved more important.
In this “full” version of a system marked by strong cleavage structures and
powerful mass-party organizations, competition between the parties was
more likely to be defensive than expansive, more oriented toward mobiliza-
tion of existing supporters than toward attraction of those already inclined to
support other parties. This point is also underlined by analyses of the specific
features of party competition in consociational democracies more generally,
and in Dutch politics in particular (Pappalardo 1981; Koole and van Praag
1990). The limitations of electoral markets in these systems were such as to
discourage party leaders from pursuing an expansive electoral strategy, with
accommodation and acceptance of mutual representation being equivalent to
“a pact among minorities who do not want and are not in a position to change
the existing distribution of power” (Pappalardo 1981: 369). In consociational
systems, dominated by mass parties, political competition was likely to be
biased towards a stable equilibrium, in the absence of which the likely out-
come was democratic collapse and/or civil war.
Even when cleavages were less deep and all-encompassing, the mass party
was often oriented more to protecting and advancing the rights and status of
its own community than to effecting any wider project for the transformation
of society. Although of course such radical transformation was often articu-
lated as the ultimate aim, particularly of socialist parties, this was coupled on

their side with recognition that they could succeed in the short run only
through the use of force, and on the side of those opposed to such transform-
ation that the threat was more hypothetical than real—with both sides accept-
ing that imposition or repression by force would be more costly than mutual
toleration.
Mutual toleration implies continuing difference. In contrast to the elite
parties that, in Burke’s terms, claimed to be pursuing the national interest,
the mass party is explicitly concerned with the pursuit or defense of the
interest of only a part of the nation, with the interests of other parts presum-
ably being pursued or defended by other parties. On the one hand, the mass
party’s status as the political arm of an already established social segment
underpins its theoretical commitment to the supremacy of the party on the
ground, served rather than dominated by the party in public office and the
party in central office, that is, to “internal party democracy.” While recog-
nizing that some tactical discretion must be left to parliamentary represen-
tatives and other elected officials who often must respond quickly to
evolving situations, on broader questions of policy and strategy, in the
mass-party view of democracy these individuals should not be independent
trustees, but rather should be instructed delegates of their parties; electors
are seen as choosing among parties (which are represented by candidates)
rather than among candidates (who happen to be members of parties), and
so whatever mandates are won belong first to the party and only secondarily
to the individuals elected. And the authoritative voice of the party is its
members, speaking either directly or more generally through the party
congress or the central office executive that the congress (or members)
elect. On the other hand, unless it is able to win an absolute majority in
parliament, the mass party’s status as the representative of only one part of
society forces it either to remain in opposition, or else to enter a coalition
with, which is to say compromise with, the representatives of other social
segments, setting up Müller and Strm’s (1999) “hard choices” among
policy (ideological purity), office, and votes.
THE CATCH-ALL PARTY
The principles that the party’s mass membership should be the ultimate
arbiter of party policy, and that the party in public office should be the
servants rather than the masters of the party on the ground, were obviously
more attractive to those on the outside of parliament looking in than they
were to those on the inside looking out. The problem for the leaders of the elite
parties as they entered the new democratic era was to mobilize mass electoral

support, and to secure provision of the greater resources required for electoral
competition with mass electorates, without giving up the independence that
they had previously enjoyed. In order to do this, they organized membership
branches like those of the mass parties, but without ceding effective control to
the party congress.
The Conservative Party in Britain offers an early textbook example. The
creation of a mass membership organization required a party in central office
to coordinate those newly organized and expanded parties on the ground, but
where in the mass party the archetypical sequence was party in central office
organizes parties on the ground in order ultimately to create a party in public
office, in the case of the pre-existing elite parties the sequence was party in
public office creates a party central office in order to organize supporters in
the form of parties on the ground.10 Nowhere is this better described and
analyzed than in the classic account by McKenzie (1963), and in particular in
his emphasis on the formal separation between the Conservative Party as
such, meaning the party in Parliament, and the National Union of Conserva-
tive and Unionist Associations, the extensive membership party on the
ground. Both were part of a party writ large, but the role of the latter was
simply advisory, or, as Winston Churchill put it, “inspirational.” Emphasiz-
ing the role of the local party worker in his address to the first National Union
Conference after the election of 1945, he argued that:
once you have an organization which has its ardent partisans in every
locality, it will be easy to build up a structure which will give these leading
local men and women an ever more effective share in inspiring the policy
of the party as a whole and help them to make a lively and vigorous
resounding contribution to the guidance which they will receive from the
summit of the Party. (Quoted by McKenzie 1963: 257)
There was another way in which transforming elite parties could not simply
adopt the mass party model whole cloth. Simply, with the possible exception
of Christian parties, the size of the social groupings that supported the parties
of the régimes censitaires were too small to sustain competitiveness on
their own.11 As a result, a strategy of encapsulation and mobilization of a
10
In Britain, the Central Office formally was formed in 1870/1, a few years after the 1867
founding of the National Union of Conservative and Unionist Associations contemporaneously
with, if not in direct response to, the suffrage expansion brought about by the Reform Act of 1867.
In practical terms, there were some local associations and some central organization before the
formal inauguration of either of these institutions. Thus the Conservatives also created the
Primrose League, founded in 1883 as a membership organization to spread Conservative
principles, but NOT as a part of the Conservative Party. In 1901 it had over 1.5 million
members with an electorate of under 8 million.
11
As an indicator of the magnitude of this problem for the parties of the régimes censitaires,
between 1895 and 1919, the proportion of the Austrian population eligible to vote increased by a

well-defined classe gardée would be recipe for disaster. Even if their old
supporters could still provide adequate material resources, the old parties
had to appeal across cleavage lines for votes if they were to remain significant
players in parliaments.
At the same time as the old elite parties were adapting organizationally to
the fact of mass suffrage, there were also changes in the nature of interparty
competition. The collapse of democracy either in fascist takeovers or by
military occupation during World War II led to reduced intensity/hostility
of conflict among the major parties after democracy was restored. As was said
of Austrian politicians, having spent the war in the same Lager (prison camp),
leaders of the Austrian People’s Party and Austrian Socialist Party were less
inclined to see each other’s subcultures (Lager in a different sense) as mortal
enemies, although with the advent of the Cold War, and frequently under
pressure from the United States, communist parties remained anathema.
Welfare states were established and later institutionalized under the guidance
of both social democratic and Christian democratic mass parties. The social
market economy was accepted in large part by both left and right, and a
reasonable consensus emerged in most democratic countries regarding the
state’s commitment to civil, political, and social rights (e.g., Scharpf and
Schmidt 2000, vol. 1, introduction).
At least in some European polities, this consensus was strongly challenged
by relatively powerful communist parties. This was another reason for the
right to live with the policy gains of the moderate left—it was a way of
protecting the capitalist economy from attacks by a more assertive left on
the fringes of the party system. Across the center at least, this meant that the
old struggles were becoming less intense. The persistent conflict over the status
of religious schools had been largely settled; there was more or less agreement
on the balance between public and private ownership within the economy;
and welfare states had become entrenched. By the 1960s, indeed, the two
major antagonists in left–right terms—the Social Democrats, on the one
hand, and the Christian Democrats or Conservatives, on the other, were
formally sharing power in Germany as well as in Austria and the Netherlands,
and were sharing “Butskellite” policies in the UK. As Kirchheimer (1957) had
earlier observed, Europe was witnessing “a waning of opposition.” Conflict,
when it occurred, looked more like a “war of maneuver” than a “war of
position,” with both sides realizing that while they might make incremental
policy gains, they also each had something to lose.
These changes in the nature of party competition were accompanied by the
beginnings of major structural changes in the society—some of which were in
factor of more than 7; in Belgium, the proportion was multiplied by 10 between 1890 and 1893, as
it was in Finland between 1900 and 1910; in Italy, it quadrupled between 1890 and 1913.

fact the result of the now more or less consensual policies associated with the
welfare state. On the one hand, social, occupational, and geographic mobility, as
well as the weakening of religious ties, helped to break down the divisions
between classes, religions, and other collective identities. Cleavage boundaries
became more porous, and cleavage communities were no longer easily isolated
from one another. To follow the Dutch metaphor from before, there was a
process of ontzuiling, or depillarization, with formerly well-bounded social
groups fragmenting and dissipating into a more individualized and particular-
ized society. On the other hand, the network of organizations in which the mass
party was embedded, and which helped to identify and mobilize the cleavage
groups, also began to fragment. The ties between mass parties and their associ-
ated interest groups began to loosen, and citizens began to turn to and organize
behind a panoply of newly independent and frequently more specialized associ-
ations, often leaving the party as such standing alone, no longer sustained by its
traditional networks of non-political organizations. Parties also lost control of,
or interest in, their own party press. Party newspapers tended to fade in import-
ance or were transformed into more independent media, and political commu-
nication was now increasingly directed towards the more neutral channels
offered by television or the commercial press.
A shared and increasingly embedded set of policy commitments, combined
with an increasingly open and depillarized electoral market, inevitably trans-
formed the strategies of the mass parties and their understandings of what
competition involved. In the first place, it made less sense for these parties to
remain tied to their original cleavage bases. These were in any case declining,
but such a strategy would also limit the parties’ capacities to reach out to the
wider and increasingly unattached electorate. Second, it made little sense for
the parties to compete on programs that advanced major social reforms or
that opposed such reforms, since in most cases reform on a grand scale was
simply not on the agenda; after the disruption and devastation of world war,
there was little appetite for potentially revolutionary (in the literal sense)
change. At most, parties might plausibly claim to change things at the
margins, or to change the manner in which a policy was managed or delivered.
Third, and following from this, it made little sense to remain tied to a
defensive or particularistic electoral strategy. Social and institutional change
may have led the once solidly rooted mass parties to become vulnerable, but it
also afforded them ample opportunity to adopt a more expansive strategy and
to compete for votes beyond their traditional constituencies.
Another strategy was clearly called for and, as Kirchheimer argued persua-
sively in the mid-1960s, it was this which led to the eventual transformation of
mass parties into catch-all parties:
The mass integration party, product of an age with harder class lines and
more sharply protruding denominational structures, is transforming itself

into a catch-all “people’s” party. Abandoning attempts at the intellectual
and moral encadrement of the masses, it is turning more fully to the
electoral scene, trying to exchange effectiveness in depth for a wider audi-
ence and more immediate electoral success. (Kirchheimer 1966: 184)
The contrast with the former mass parties was marked. These new parties were
much less easy to differentiate from one another, and their programs tended to
crowd in the center, offering far less that was distinctive or particular. They
found themselves appealing for support in the same newspaper columns and
television studios. Although parties remained publicly identified as the repre-
sentatives of particular interests, at least across the mainstream parties of all
types were becoming increasingly willing to accommodate the interests voiced
by others as well.
As already suggested, a strategy of trying to appeal across segmental—and
especially across class—boundaries was nothing new for the transforming
elite parties; although the self-identified “middle class” would expand enor-
mously in the second half of the twentieth century, particularly in the early
days of mass suffrage a party that could attract only the votes of the middle
(and upper) classes was doomed to defeat. Instead, the transforming elite
parties retained the idea of an overriding and singular national interest, and
in contrast to the idea of social segmentation characteristic of the mass party
(and the associated idea of “real workers” representing the working class, for
example), had appealed to a more traditional and hierarchical sense of society,
and the presumed desire of citizens to be governed well more than and before
they desired self-government.12 With the “big” issues apparently settled, with
smaller shares of the electorate to be mobilized by feelings of solidarity and
larger shares to be attracted by appeals to immediate interests, and, especially
for the leaders of the mass party in public office, with the now real possibility of
enjoying government office if sufficient votes could be secured, appeals on bases
other than group solidarity that had characterized elite parties since suffrage
expansion, and the associated movement in the direction of “catch-all-ism,”
made sense for the leaders of the mass parties as well.
The transformation of mass parties into catch-all parties was not, however,
as unproblematic as this description of social trends and political incentives
may suggest. While the force of ideology may have been waning, at least in
the early part of the post-war period, ideology remained significant, with
Clause IV (“To secure for the workers by hand or by brain the full fruits
of their industry and the most equitable distribution thereof that may be
possible upon the basis of the common ownership of the means of produc-
tion, distribution and exchange, and the best obtainable system of popular
12
This appeal is epitomized. for example, in the phenomenon of the “working-class Tory”
(Nordlinger 1967).

administration and control of each industry or service”) still firmly in the
constitution of the British Labour Party and the 1959 Godesberger Programm
in which the German Social Democratic Party renounced Marxism and
accepted that “Private ownership of the means of production can claim
protection by society as long as it does not hinder the establishment of social
justice” still in the future.
Unsurprisingly, attachment to ideology tended to be strongest among the
activists of the party on the ground. On the one hand, “commitment to the
cause” often was instrumental to their becoming activists in the first place, and
thus in distinguishing them from ordinary party members, not to mention
supporters who had not even taken the step of joining the party. On the other
hand, they did not personally confront the constraints, or personally share in
the material and social rewards, of being in office. For the activists, the
tradeoff between maintaining ideological purity versus compromising in
order to achieve office was weighted more heavily in favor of the first, while
for the party in public office it was weighted more heavily in favor of the
second. Analogous to the desire of the party in public office of the old elite
parties to establish membership organizations that would support, but not
constrain them, the party in public office of the mass parties needed to loosen
the constraints imposed by their pre-existing membership organizations if
they were to pursue the increasingly necessary catch-all strategy in the elect-
oral arena, and even more if they were to govern effectively. Thus we return to
McKenzie’s finding that while the parliamentary Labour Party might be
subservient to the party’s National Executive Committee when in opposition,
once in government, the Labour Party was dominated by its party in public
office in the same way as the Conservative Party was dominated by its party in
public office. Along with increasing similarity of policies and strategies, the
old elite and the old mass parties grew increasingly similar to one another
organizationally as well.
While the intention of the leaders of the old elite parties may have been that
their parties on the ground would be no more than organized cheerleaders for
the professional politicians in the party in public office, it did not always work
out like that in practice. Once recruited, party members started to make
demands, abetted by the principle that had been first articulated as part of
the ideology of the mass party, but then had been more widely accepted, that
the party in public office should be responsible to the party’s members. The
result was that although the party in public office might have been more
dominant in the transforming elite party than in the case of the mass party,
its dominance was constantly under challenge.
This challenge was furthered by the same social changes that were under-
mining the mass party. Reduced working hours, increased, and increasingly
standardized education, the political eclipse of the traditional upper class,
and, indeed, a general weakening of class divisions, all combined to make

expectations of deference to party leaders more problematic. Rather than
owing their positions as party leaders to their positions at the top of a general
and natural social hierarchy, leaders in the transformed elite parties, like
leaders in other areas of community life, increasingly had to justify their
leadership positions with reference to their capacity to satisfy the needs of
their followers, and the followers increasingly had the capacity and the
inclination to define and articulate those needs for themselves.
The mass party tended to arrive at a similar result, but it came there from
the other direction (see also Svåsand 1994)—that is, through the increased
assertiveness of the party in public office rather than the increased assertive-
ness of the party on the ground. Once significant influence over government
policy and entry into government office were perceived to be realistic possi-
bilities, the leaders of mass parties (particularly those in the party in public
office, but often those in the party in central office as well) tended increasingly
to orient themselves toward the requirements of electoral victory, and to be
constrained by the realities of governing. The result from either perspective
was to exacerbate tension between the party in public office and the party
on the ground, a tension that was to become a persistent feature of the catch-
all party.
The weakening, and in some cases the outright dissolution, of the organ-
izational ties between parties and their ancillary interest organizations, and
the increasing prominence of non-party organizations promoting specific
interests without reference to broader patterns of (in any case weakening)
social segmentation, also led to a repositioning of parties in the wider political
process. Rather than each representing (negotiating on behalf of) specific
interests that were part of its own political family (a strategy that
LaPalombara (1964), writing from the perspective of interest groups rather
than parties, described as parentela), as called for by the newly prominent
pluralist model of democracy (Truman 1951; Dahl 1956) all of the parties
tended to act as brokers among competing interests, and between those
interests and the state. Moreover, having loosened or shed ties to individual
parties, the groups representing the various interests came to expect their
concerns to be heard and taken seriously by all governments, regardless of
their partisan complexion. Although one party might be biased towards some
interests while another party might be biased towards others, in their common
role as brokers rather than purely as representatives, the parties came to be
more similar to one another.
In their essentials, mass parties and catch-all parties can be contrasted on
two crucial dimensions: on the one hand, in terms of the nature of electoral
competition in which they were engaged; on the other, in terms of the
profundity of the social and economic stakes for which they competed (see
Figure 2.1). The mass party seeks to mobilize its own supporters within its
own cleavage heartland, whether this be defined in class terms, or religious
Competitive mobilization of Mobilization of party

party classes gardées identifiers and competition
for the “floating” vote
Low stakes: For the left, Catch-all party

amelioration and reform. For
the right, gradual adaptation.
Veto capacity used to
strengthen bargaining
position
High stakes: For the left, Mass party

societal transformation. For
the right, prevention of social
revolution. Veto capacity
used to thwart the other side
F I G U R E 2 . 1 The mass party and the catch-all party
terms, or whatever. For this it requires committed party members and activists
on the ground, as well as members and adherents of the myriad of other
organizations and associations associated with the cleavage group—unions,
churches, sports clubs, and so on. In this war of position, what matters most
is the share of one’s own cleavage group that could be brought to the polls. The
stakes in this enduring battle were also relatively high—at least initially—
involving social transformation (from the point of view of the left) or resistance
to such transformation (from the point of view of the right and religious
groups). Both sides, through mobilization, also sought to act as veto players
with the objective of being able to block unpalatable reforms even when
confined to opposition. The catch-all party, by contrast, competes for relatively
low stakes, being content to ameliorate or reform a given policy program, or to
maintain or adapt an established public commitment. To the extent that it has
the capacity to act as a veto player, it tends to use that power as a bargaining
chip to exact compromise rather than reflexively blocking proposals that it finds
unpalatable. It also competes as an effective broker of different interests,
claiming to be the more effective representative of any one of a set of competing
electoral claims. The party thereby moves away from its classe gardée, seeking
and winning votes wherever they might be found. There are no great struggles
involved in such competition, and no enduring interests at stake. Instead, any
lingering ideological and purposive commitments take second place to the need
to win office and votes.
That said, there is one crucial feature that is common to both types of party,
and that distinguishes both from the form of party that began to emerge in
Europe in the last years of the twentieth century. That is, both the mass party
and the catch-all party aimed to represent (or broker) established interests. In

the case of the mass party, these interests were a fixed and enduring part of the
cleavage subculture—the socialist trade unions, the Roman Catholic church,
the association of smallholders, and so on. In other words, the interests within
the society were divided and allocated among the parties in a system of
enduring ties. In the case of the catch-all party, by contrast, the interests
competed in the market place, and belonged to no one party in particular.
Some unions, some religious groups, or some farmers might shift to one party
or shift to another, but their relationship to each of the parties was contingent
and instrumental. In the mass party, interests belonged to the party subcul-
ture. In the case of the catch-all party, interests competed and were brokered.
In both cases, however, politics continued to revolve around the represen-
tation of interests—whether this were in the electoral, parliamentary, or
governing arenas.
It is our contention that both of these models are now largely defunct. This
does not mean that they are only of historical interest, however. Rather—and
this is partly to do with how politics itself developed, and partly with how the
political science profession developed—they are also of interest politically,
normatively, and scientifically, in that they have set the terms of reference for
most post-war party studies, as well for many political activists and theor-
ists.13 One symptom is a near universal consensus that the mass party is the
ideal model for parties and political leaders to follow, and that all subsequent
variants, including the catch-all party, are therefore a disappointment.
Indeed, Kirchheimer himself was among the most explicit in expressing his
regret at these developments, arguing that the citizen found in the catch-all
party “a relatively remote, at times quasi-official and alien structure” (1966:
199). The transition from mass party to catch-all party, he argued, involved
13
In particular, the mass party ideal of an internally democratic organization ultimately
responsible to its members is regularly reflected in pronouncements of organizations aiming to
promote democracy around the world. When the European Commission for Democracy through
Law (the “Venice Commission”) issued its Code of Good Practice in the Field of Political Parties,
it identified “to reinforce political parties’ internal democracy” as “its explicit aim.” Likewise, the
Parliamentary Assembly of the Council of Europe has called on member states to “ensure that the
legislative framework promotes the implementation by political parties of internal party
democracy principles.” After a workshop on the subject at the Third Assembly of the World
Movement for Democracy, “There was a general consensus at the workshop that the
strengthening of internal party democracy is a crucial prerequisite for democratic development
in various countries.” International IDEA has a project on internal party democracy that aims “to
provoke party reform by identifying the challenges facing political parties for them to become
more democratic, transparent and effective.” According to the Netherlands Institute for
Multiparty Democracy, “internal democracy” is one of the “ ‘institutional guarantees’
that . . . political parties would have to fulfil if they were to effectively meet what is expected of
them in a democracy.” Similarly, the United States Agency for International Development’s
support for political parties “emphasizes the need for internal party democracy.”

the party’s transformation from an organization combining the defense of
social position, the quality of spiritual shelter, and the vision of things to
come into that of a vehicle for short-range and interstitial political choice
[that] exposes the party to the hazards of all purveyors of nondurable
consumer goods: competition with a more attractively packaged brand of
a nearly identical merchandise. (1966: 195)
In other words, while the changes that led to the decline of the mass party and
the emergence of its successors may have resulted from rational decisions on
the part of party leaders and members, and while they may well have been
rendered more or less inevitable as a consequence of changes in the social and
economic environments within which parties competed, they have neverthe-
less contributed to creating a sharp disjuncture between democracy as it has
been practiced since at least the 1960s, and democracy as it is usually
understood—and, as in Kirchheimer’s case, probably wished for—in the
traditional models of party government and party democracy, that were
based on the idea of party as the political expression of well-defined political
community and articulating a relatively stable and overarching political
philosophy.
The Locus of Power in Parties
Each of the party models that we discussed in Chapter 2 is associated with a

different balance between the three faces of party organization—the party in
public office, the party in central office, and the party on the ground. In the
elite party model, the party in public office is clearly the dominant face—
indeed, it is effectively fused with the party on the ground in a system in which
there is no real need for a central office. In the mass party model, it is the party
on the ground that is clearly dominant in principle, while it is often the party
in central office, the organizer and representative of the party on the ground,
that is dominant in practice. In any case, the nexus between the party on the
ground and the party central office is often underlined by their shared incen-
tive to keep the party in public office in a subordinate position. Seen from
this perspective, the essence of the catch-all party is that it is marked by an
ongoing tension or conflict among the three internal faces. The place in which
this conflict is played out is the party central office, and the key question is
whether the central office will act as the agent of the party on the ground
in controlling the party in public office, or rather as the agent of the party in
public office in organizing and directing its supporters on the ground.
The party in central office developed in tandem with the party on the
ground. Once a party began to nurture a membership organization, and
once it began to attract and potentially integrate large numbers of more active
voters and militants within its own immediate ranks or within its broader
penumbra of supporters and associated organizations, it required personnel,
offices, and institutions that could maintain and manage this membership and
act as a conduit between the party on the ground and the party in public office.
As with any large-scale membership organization, some form of central
coordinating office was required.
Once a central office develops, however, it—and through it, the
membership—begins to curtail the independence of the party in public office,
albeit in a form which, following Michels, can disproportionately favor the
party in central office at the expense of the party on the ground. In an attempt
to reassert its power, and to lessen the constraints imposed on it by the wider
party organization, the party in public office in the catch-all party might
sometimes attempt to cut itself free from the party on the ground, insisting
on leadership flexibility and denying the membership a voice in key strategic
and policy decisions; or it might itself seek to pack the party in central office,

frequently with its own leaders as ex officio members of the party’s central
committees, who could then dominate the appointment of central office
personnel, and thus wrest control of the party as a whole from the hands of
the party membership and its bureaucracy. This ushers in a more or less
permanent conflict among the three faces, with the competition between the
party on the ground and the party in public office for influence within the
party and for control of the party in central office becoming a central element
within the catch-all party model.
In this chapter we contend that the development of party organizations has
now gone beyond the catch-all period and has entered a new phase, in which
parties have become increasingly dominated by, as well as most clearly
epitomized by, the party in public office (Katz and Mair 2002).1 In consider-
ing the relationship among the extraparliamentary organizations (the party
on the ground and central office in our terms), the parliamentary party
groups, and the party in government (the latter two both in our party in
public office) in European democracies, for example, Heidar and Koole
(2000: 265) see both that the frequency of conflict among these three faces
had decreased in the 1980s and 1990s, and that the influence of the parlia-
mentary party groups had increased. Even as some parties try to maintain an
appearance of independence of the central office and party in public office, for
example by making ministerial office incompatible with holding the position
of party president or secretary, the incumbents of these party offices are
often members of the party in public office and “ministers in waiting,” as
illustrated by Elio Di Rupo who stepped down as chairman of the Belgian
Parti Socialiste in 2011 when he became prime minister, but resumed the
chairmanship in 2014 when liberal Charles Michel became prime minister;
during much of his time as party chairman, Di Rupo was also minister-
president of the Walloon region, and clearly a leader in the party in public
office as well as being head of the party central office.2
The first and most obvious symptom of this new pattern in the internal
balance of power involves the distribution of financial resources within the
party, and, in particular, the distribution of state subventions. Since the 1960s,
1
For example, in discussing the fact that Dutch ministers must resign their seats in Parliament
(Constitution of the Netherlands 2008, Art. 57), Andeweg and Irwin (2002: 141) observe that
“whether the party was unable to decide whether its leader in government or its leader in
Parliament was to assume the overall leadership of the party . . . the latter eventually emerged
victorious” without even considering the leader of the party on the ground or the central office.
2
See also the example of Yves Leterme who resigned as president of the CD&V to become
minister-president of Flanders (Pilet and Wauters 2014: 44). The British Labour Party’s rules
make the position of general secretary incompatible with candidacy for, let alone membership of,
parliament, but the rules justify this on the ground that the general secretary should “devote her or
his whole time to the work of the Party,” rather than a desire to maintain a separation between the
central office and the party in public office (Labour Party Rules 2017, chapter 4, clause II.4.A).
The Locus of Power in Parties 55

when direct state subsidies to political parties were first introduced in a limited
number of countries, the channeling of state aid to party organizations has
become an almost universal practice in modern democracies. In most coun-
tries, these subventions were first allocated to the parliamentary fractions of
the parties, and only later was the practice extended to include direct subsidies
to the central party organization itself. Even when funds were extended
beyond the parliamentary and governing offices of the party, however, it
was usually only a smaller proportion of the subsidy that was involved,
although by now many party head offices pay as much as half of their budgets
from public subsidies (Nassmacher 2009: 333; Katz and Mair 1992c).
Precisely who within the party leadership decides how these sums are then
allocated across items within the parties’ own budgets is, of course, not easily
known, and in this sense the mere existence of subsidies may not seem a strong
indication of the privileging of the party in public office—but even less is a
large central office budget evidence against such privileging unless one knows
who controls the central office. But the fact that the process of state subven-
tion was often initially limited to the parliamentary fractions of the parties,
that the fractions themselves often continue to receive the greater share of the
total subsidy, and that it is in parliament that the final decisions are taken as to
the levels and types of subsidy to be made available, all suggest that the
increasing availability of state aid is one of the factors operating to the
potential advantage of those in control of public office—and in any case by
lessening the relative importance of membership payments, state aid also
reduces the dependence of both the party central office and the party in public
office on the party on the ground, and to the extent that “he who pays the
piper calls the tune” it thereby lessens the relative importance of the party on
the ground in its competition with the party in public office for control over
the central office.
The second symptom, and partly a consequence of the availability of state
subsidies, is that by the end of the 1980s a clear shift had begun to take place
within party organizations in the allocation of party staff. Such comparable
time series data on party staffs as are available (Katz and Mair 1994; Bardi
et al. 2017) contain clear evidence of a common trend across countries and
parties for the numbers of staff employed by the parliamentary parties, and
hence by the party in public office, to significantly outstrip the numbers
employed by the party headquarters. Indeed, across all the countries for
which comparable data are available over time, the average balance shifted
from somewhat more than 25 percent of staff being employed within the
parliamentary offices in earlier periods (usually in the 1960s or early 1970s)
to slightly more than 50 percent by the late 1980s. Although in some countries
this shift was very substantial, and in other countries almost negligible, there
was by the late 1980s no single country that defied this general trend, and such
limited evidence as does exist suggests that this trend continued into the 1990s

(Webb et al. 2002; Bardi et al. 2017). Given that staff levels constitute a crucial
organizational resource, these data also confirm an increasing bias in favor of
the party in public office.
The third symptom that is relevant here (see for instance Katz and Mair
1995) is that most substantial and/or enduring parties in the long-established
democracies have enjoyed periods of office in national governments and now
orient themselves as a matter of course to the occupation of public office.
Although Italy was an extreme case, in 1983 slightly over 40 percent of the
vote went to parties that were effectively outside of contention for govern-
ment. In the early 1970s (assessed at the first general election after 1970),
Belgium, Denmark, Italy, Luxembourg, the Netherlands, Norway, Sweden,
Switzerland, and the UK all saw at least 14 percent of the vote going to parties
that had not been in national government at any time in the preceding twenty
years. Assessed somewhat differently, in the period between 1961 and 1971,
Denmark, Finland, Germany, Ireland, Italy, Luxembourg, Norway, Sweden,
and the UK all had at least one party that won at least 10 percent of the vote
(and so by most standards would qualify as significant) that had not been in
the national government for at least fifteen years (long enough to think of
themselves more as parties of opposition than as potential governors).3 By the
last decade of the twentieth century, however, there were few, if any, signifi-
cant parties of enduring opposition in the long-established democracies, while
almost no such party could be found in the new democracies of post-
communist Europe; in the Italian case, over 96 percent of the vote in the
1994 general election went to parties that were included in one of the three
alliances seriously contending for national office, and after Rifondazione
Comuninsta entered the Prodi government in 1996, the share of the electorate
that had voted for parties never in government fell to below 5 percent. Such
parties as were excluded from government office, or which excluded them-
selves, were generally small and either at the extremes of the dominant
(generally economic left/right) axis of competition, or else represented minor-
ity ethnic, regionalist, or environmental demands. The mainstream parties, on
the other hand, including a substantial number of Green parties and succes-
sors of communist parties that had reconstituted themselves as mainstream
social democrats, as well as some of the representatives of the far right, could
3
The parties are: Denmark: SF, 10.9% and 20 years in 1966; Finland: SKDL, 22% and
16 years in 1962; Germany: SPD 39% and 19 years in 1965; Ireland: Lab 17% and FG 34.1% and
15 years in 1969; Italy: PSI 13.8% and PCI 25.3% and 17 years in 1963; PCI 27.9% and 22 years in
1968; Luxembourg: KPL 15.5% and 22 years in 1968; Norway: H 20% and 15 years in 1961;
Sweden: Lib 14.3% and M 12.9% and 22 years in 1968; 16.2% and 11.5% and 24 years in 1970;
UK: Liberals 11.1% and 18 years in 1964. Note this excludes Belgium (RW and VU) and
Denmark (Progress) for which the non-governmental parties were not themselves at least
15 years old.

all be considered to be conventional holders of government positions. Of the
nine countries with parties that won 10 percent of the vote and had been out
of government for at least fifteen years in the 1961–71 period, only Italy,
Luxembourg, Norway, Sweden, and the UK remained in the 1990s—and
Italy (as noted above) and Luxembourg were both out of this group by the end
of the 1990s. In contrast to the interwar and immediate post-war decades, by
2000 there were virtually no significant parties of long-term opposition.4
This development has almost certainly had a significant impact on the
internal balance of organizational forces within the parties concerned, and
Panebianco (1988: 69) is not alone in pointing out that “the organizational
characteristics of parties which are in opposition for a good part of their
existence are different from those which stay in power for a long time.”
Indeed, we suggest that much of the organizational restyling of parties in
recent years has been heavily influenced by the degree of commitment to and
involvement in the governmental process. With governing becoming a stand-
ard experience and expectation for most parties, there has been an enhance-
ment in the status, prestige, and autonomy of the party in public office. What
we see is not only a process of the “parliamentarization” of parties (Koole
1994: 291–2), but also one of “governmentalization” (Müller 1994: 73), a
trend that inevitably reduces the relative importance of both the party on the
ground and the party in central office.
One important consequence of this has been the tendency to increase the
degree of representation and, presumably, the degree of influence, afforded to
the party in public office within the committees of the party in central office.
Parliamentarians and their leaders now tend to be accorded greater weight in
these bodies than was the case in the 1960s and 1970s, and correspondingly less
weight is given to the otherwise non-office-holding representatives of the party
on the ground. Though not universal, this trend is sufficiently common to
imply that, more often than not, the party in public office now exerts substan-
tially more control over the national executive than used to be the case.
The exceptions to this pattern are also interesting, however. In her study of
the early development of party organizations in the new Southern and East
European democracies, for example, Ingrid van Biezen (2003) found that
despite much overlap in personnel between the party in public office and
party in central office, and despite the substantial powers of nomination
from the former to the latter, it was the central office that was emerging as
the dominant actor. In part, she suggested, this reflected the sheer newness of
the democracies and of the parties themselves, marking “a desire to increase
party cohesion and so reduce the potentially destabilizing consequences of
4
By the first decades of the twenty-first century, this situation had changed again, with a
dramatic increase in electoral support for new parties opposed to all of the mainstream parties as a
group. We consider this development in Chapter 7.

emerging intraparty conflicts, which themselves are an inevitable by-product
of the context of weakly developed party loyalties and a generalized lack of
party institutionalization.” The party in central office also offered the new
party leaders a more secure foundation, in that by basing themselves within
the party executive, they take advantage of “a relatively stable and predictable
organizational foundation, which is a particularly valuable asset in a climate
of frequent party ruptures” (van Biezen 2003: 218). In none of these cases,
however, was the party in central office acting as the agent of the party on the
ground. On the contrary, the party in central office simply offered the advan-
tage of a stable and secure foundation from which party leaders—the party
top—could manage the party as a whole.
Another exception was Belgium, where parties experience severe difficulties
in seeking to coordinate their presence in a variety of multi-level public offices
and where overall control of the various fractions of the party has been
facilitated by incorporating the leaders of these fractions into a single party
executive. In its standing orders adopted in 2013, for example, the Franco-
phone Belgian Socialist Party makes its members of European Parliament or
Commission, members of parliament and ministers at the federal, regional, and
language community levels all ex officio members of its Comité fédéral, subject
only to the proviso that the ex officio members cannot represent more than 49
percent of the votes on the committee. Deschouwer (1994: 94) interprets the
evidence of overlapping personnel as a sign of the central party seeking to
control and coordinate the various parties in the various public offices rather
than of the party in public office dominating the party in central office, but in this
it also underlines the difficulty of trying to assess the relative influence of two
institutions that are to a great degree staffed and led by the same people. On the
other hand, the fact that the position of regional minister-president, let alone
the prime ministership of the country, appears to be regarded as superior to
the party presidency suggests that the central office is a place for coordinating
rather than directing the multiple parties in public office.
A more recent impression of the situation in these countries can be taken
from the Political Party Database (PPDB) assembled by Scarrow et al. (2017).
These data code the name of the leader of the parliamentary party, the de
facto leader of the national party extralegislative organization, and “the most
important political leader of the national party.” Table 3.1 shows the patterns
of overlap for the 34 data points from Belgium plus the new democracies
included in the PPDB. The most obvious point is that there were no cases in
any of these systems in which the parliamentary and extraparliamentary
leaders were different and yet the parliamentary leader was coded as the
more important. While this might suggest a secondary position for the party
in public office, such a conclusion would be quite misleading. In many cases,
the “most important” leaders appear to have owed their importance to
occupancy of ministerial office rather than to their party positions (although

T A B L E 3 . 1 Party leadership in Belgium, the Czech Republic, Poland, Portugal, and Spain
Country Parliamentary party and Parliamentary party and extralegislative N

extralegislative leadership leadership not held by same person;
held by same person extralegislative leader is “most important”
Belgium 0 9 9
Czech Republic 5 0 5
Poland 2 4 6
Portugal 0 7 7
Spain 5 2 7
Source: Scarrow et al. (2017)
in the context of a full political career it may be difficult to say which has
either temporal or substantive primacy), and almost all were members of the
national (or in the case of Belgium and Spain regional) parliaments, and thus
were members of the party in public office. Even here, in other words, those
who hold public office appear to be in a dominant position; the only question
is whether they exercise that dominance over or through the central office.
In other polities, where pre-existing loyalty to the party itself (Bowler et al.
1999: 7) among MPs could be assumed, a secure foundation was provided by
the party in public office, with the party in parliament in particular constitut-
ing a relatively stable and increasingly well-resourced base from which to
direct party affairs. This is not least because, as noted above, the growth in
organizational resources, as indicated by staff and money, has tended to be to
the advantage of the parliamentary party, while those resources that remain
within the central office tend nowadays to be devoted to the employment of
contractual staff and consultants, and to the recruitment of outside expertise
(Webb et al. 2002). In such a context, professional capacity matters more than
political accountability, and this development also undermines the independ-
ent political weight of the party central offices. It is interesting to note, for
example, that while it often proves very difficult to identify the electoral
impact of the development of new campaign techniques and technologies, it
is nevertheless clear that they have helped to shift the weight of influence
within party organizations from amateur democrats to the professional con-
sultants who control these techniques (Bartels 1992, 261; see also Panebianco
1988, 231–2).
More specifically, the gradual replacement of general party bureaucrats by
professional specialists acts to “depoliticize” the party organization and to
create the conditions within which the leadership in public office wins more
autonomy. This is particularly so in that the activities of the new professionals
are almost always more directed (externally) at winning support within the
electorate at large rather than (internally) at nurturing the organization and
maintenance of the party on the ground. In Plasser and Plasser’s (2002) survey

of campaign professionals, for example, it was noted that the external
consultants, who increasingly dominate campaign activities and planning,
tend to speak of the role of the party organizations and—by implication—
that of the party central offices “in sceptical and distant terms, pointing in the
direction of an adversarial relationship between highly specialized campaign
professionals and the traditional political party organization” (2002: 316).
Indeed, fewer than one third of those interviewed in the survey in the West
European region regarded “a strong and effective party organization” as
being a valuable resource in determining electoral success (2002: 317–18).
Moreover, unlike in the United States, where each political consultant tends
to be associated with only one of the parties, in some other countries they may
shift allegiance between parties, just as an advertising agency may move from
advertising one brand of automobile or beer or soap powder to another. In the
UK, for example, the firm of Saatchi & Saatchi, which was central to the
success of the Conservative Party campaigns in 1979, 1983, and 1987, was in
2007 hired by the Labour Party.
This also underlines a further important shift in the general orientation of
modern party organizations. As television and the mass media more gener-
ally, not to mention the internet, have emerged as the key channels of
communication between party leaders and voters, offering the benefits of a
direct linkage in place of what previously had been mediated by organiza-
tional cadres and activists, party campaigning has become more centralized
and “nationalized,” with the core of the parties’ messages now tending to
emanate in a top-down fashion from a single national source. In most polities,
a specifically local input has become less and less relevant to the national
campaigns, and this also implies that the parties at the national level need to
devote less and less effort to the organization and mobilization of the party on
the ground. Resources become devoted instead to selling the party message to
the electorate at large. Moreover, the party message itself has come to be
increasingly identified with, if not simply defined by, the personality of the
party’s leader who is almost always based in the party in public office.5 The
result is not only a changed—and more professionalized—role for the party
central office, but also the eventual erosion of the division regarding respon-
sibilities between the party apparatus in central office and that in public office.
5
See, for example, Table 4.7. Although personalization of party messages appears to have
increased in recent decades, it is by no means new. See for example the German CDU poster from
1957 that showed the face of Konrad Adenauer and the phrase “Keine Experimente!” (no
experiments) (https://www.google.com/search?q=adenauer+keine+experimente+poster&tbm=isch
&source=iu&pf=m&ictx=1&fir=c2TrJKQPWehkpM%253A%252Cl2VDceSzp8uWbM%252C_
&usg=—UhaYHObiXGs_Du8fP6BvJOSNwkg%3D&sa=X&ved=0ahUKEwiRzeCf9pj
XAhUL7CYKHbQNAX cQ9QEIMTAC#imgrc=c2TrJKQPWehkpM:).

As parties become more externally oriented, the roles of the professionals
serving the party in central office and of those serving the party in public office
become almost indistinguishable, with both sets responding in the main to the
demands of the party leadership in parliament and in government. This is also
reflected in the shift of intraorganizational party power to the office of the
leader, a trend documented by Poguntke and Webb (2005: chapter 15; see also
Gauja 2013 with particular regard to autonomy with regard to policy), who
see in this process widespread evidence of the “presidentialization” of con-
temporary European democracies and thereby of the weakening of parties as
collective actors. Not only have political leaders in government cumulated
executive power within their personal offices, but they have also cumulated
party power. As they say (2005: 343–4):
Developments in the party face of presidentialization can be reported very
straightforwardly: none of the country experts doubts that the leaders of
(potentially) governing parties have enjoyed a growth in intra-party power
and/or autonomy, or these were already comparatively high at the start of
the period analysed and have remained so, in each and every case.
THE EROSION OF THE PARTY ON THE GROUND
All of this might well lead to the expectation that, perhaps with a few
exceptions, the party on the ground in modern mainstream parties would
wither away, or at most remain as a vestigial organ, with no essential function,
and with few if any resources devoted to its maintenance. With the party
central office increasingly staffed by professionals who provide the (elec-
tronic) conduits for communication between the party and the electorate
that used to be prominent among the raisons d’être of the party on the ground,
and who increasingly are paid from state subventions or large donations,
neither of which require an extensive membership organization, the party
on the ground would be marginalized as an independent force within the
party writ large. Seen from this perspective, the leaders would increasingly
become the party; and the party would become little more than the leaders.6
In reality, it is easy to overstate this case, just as in the past it was easy
to overstate the importance of the party on the ground in defining and
controlling the mass party. Nonetheless there are unmistakable trends in
this direction.
6
Particularly with regard to American parties, this has always been the position of those who,
like Downs, Schumpeter, and Schlesinger, simply identify parties as teams of political leaders.

The most obvious symptom of such a change is, of course, the physical
withering of the party on the ground, with available evidence pointing to a
gradual but almost universal decline in the sheer size of party memberships.
The relevant data, drawn from Katz et al. (1992) and van Biezen et al.
(2012), are summarized in Table 3.2, which reports levels of party member-
ship as a percentage of the electorate in nineteen democracies in contempor-
ary Europe. Although comparisons over the full stretch of time are limited,
in that comparable data are available only for ten countries in the 1960s, the
trend is readily apparent. Party membership as a share of the electorate
averaged close to 15 percent at the beginning of the 1960s, and then fell, in
that same group of polities, to 10.7 percent around 1980, to 9.2 percent
around 1990, and then to 6.3 and 5.6 percent around 2000 and 2010. By the
end of this period, in other words, levels of party membership had shrunk by
almost two thirds. The same pattern is evident when we look at larger groups
of countries over shorter periods. In the group of sixteen democracies for
which data are available from the 1980s onwards, the percentage goes from
8.6 percent to 7.6 percent around 1990, and then to 5.6 and 4.9 percent in the
latter decades. Even across a shorter period, membership drops by almost
one half. Partially updating these figures, Webb and Keith (2017: 33) find
total membership as a proportion of the national electorate to have declined
from the latest figures reported by van Biezen in fourteen of the fifteen
countries for which they can make a comparison. (The exception is Ireland,
although the raw number of members also increased slightly in the Netherlands
and the UK as well.) While the trends are not all strictly monotonic, only
Greece, Spain, and Slovakia defy the general trend of decline, and these were
all new democracies.7
Treating membership as a percentage of the electorate is the most effective
way to compare trends over time and across countries, but as with any
proportion, percentages can decline as a result of an increasing denominator
(a larger electorate) as well as a shrinking numerator (fewer members). This is
controlled in Table 3.3, which reports changes in the absolute numbers of
members in the sixteen democracies for which data from the 1980s are
available, and here too the evidence is unequivocal. In all but three of the
countries—Greece, Portugal, and Spain, each democratizing in the mid-
1970s—there has been a major decline in the absolute numbers of party
members. Even in Germany, numbers of members have fallen substantially
despite the potential for growth in membership numbers resulting from the
7
As Scarrow (2015) points out, some of the apparent decline may be the result of changing
definitions of membership (e.g., the decisions of Norwegian and Swedish labor parties to stop
counting indirect members in the 1990s) or improved record keeping that purged non-existent
members from the roles. Nonetheless, while these may reduce the magnitude of the decline, they
do not alter the direction or near universality of the trend.

T A B L E 3 . 2 Party membership as a percentage of the electorate, 1960–2010
Approximate dates
1960 1970 1980 1990 2000 2010
Austria 26.2 28.5 23.7 17.7 17.3

Belgium 7.8 9.0 9.2 6.6 5.5
Denmark 21.1 7.3 5.9 5.1 4.1
Finland 18.9 15.7 13.5 9.7 8.1
Germany (West Germany 2.5 4.5 3.9 2.9 2.3
through 1990)
Italy 12.7 9.7 9.1 4.1 5.8
Netherlands 9.4 4.3 3.2 2.5 2.5
Norway 15.5 15.4 13.1 7.3 5.0
Sweden 22.0 8.4 8.0 5.5 3.9
UK 9.4 4.1 2.6 1.9 1.2
France 5.1 3.0 1.6 1.9
Greece 3.2 6.3 6.8 6.6
Ireland 5.0 4.9 3.1 2.0
Portugal 4.9 5.1 4.4 3.9
Spain 1.2 2.1 3.4 4.4
Switzerland 10.7 8.0 6.4 4.8
Czech Republic 7.0 3.4 2.0
Hungary 2.1 2.2 1.5
Slovakia 3.3 4.1 2.0
Source: Katz et al. (1992); van Biezen et al. (2012)
T A B L E 3 . 3 Change in the numbers of party members 1980–2010
Country Change in the number of Change in numbers as % of

members original membership
Austria 422,661 28.6

Belgium 191,133 31.0
Denmark 109,467 39.7
Finland 260,261 42.9
France 923,788 53.2
Germany 531,856 27.2
Ireland 50,586 44.7
Italy 1,450,623 35.6
Netherlands 126, 459 29.4
Norway 288,554 62.6
Sweden 241,130 47.6
Switzerland 178,000 43.2
UK 1,158,492 68.4
Greece + 335,000 + 148.9
Portugal + 4,306 + 1.4
Spain + 1,208,258 + 374.6
Source: van Biezen et al. (2012)

incorporation of the former East German Länder into the Federal Republic
in 1990. When we take all the countries together, even including the three
newer democracies, we see almost 4.5 million fewer members in 2010 than
there were in 1980.
Party members are not only less numerous, but also have changed in
character. Survey data suggest that they are now more moderate in ideo-
logical terms than were the larger memberships of the 1960s and 1970s, and in
this sense more representative of the voters. But they are also less representa-
tive in social and economic terms, being increasingly male, better-off, and
older than their counterparts from earlier decades (see Scarrow and Gezgor
2010). They also tend to be more narrowly drawn in occupational and status
terms than the earlier memberships. Data from the European Social Survey
indicate that party membership is more likely to be drawn from among those
working in the public sector than those in the private sector (van Biezen et al.
2012), while more detailed surveys of party members themselves indicate a
similar bias. In the case of the Labour Party in Britain, for example, Seyd and
Whiteley (2002: 40–5) found a much higher proportion of occupations from
the public and voluntary sectors among the membership than among the
party’s voters. The same pattern was caught in the Dutch case, where 23
percent of the active members across all parties (and 17 percent of the non-
active members) were employed in the public sector, as against some 12
percent of voters (Koole and van Holsteyn 1999: table 1). This bias towards
the public sector, and hence the greater potential for a sharing of interests
between members and their administrative as well as political superiors,
suggests that there may be a much closer link between the membership and
the political leadership than is implied by the simple “three faces” approach.
Indeed, when viewed in terms of profile, character, and sheer size, it suggests
that membership may be becoming less a genuine party on the ground, but
instead is transforming into something that is akin to the outer ring of the
party in public office.
In this regard, it is also interesting to note the evidence of professional
ambition in Bruter and Harrison’s (2009) data on young party members. In
their six-nation survey, they find that young party members break down into
three distinct groups—the moral-minded members, who emphasize the
importance of good citizenship and helping others; the social-minded mem-
bers, who emphasize the importance of solidary incentives and interaction
with others; and the professionally minded members, who emphasize “a desire
to achieve positions and honors, become a politician, and derive money or
material benefits from party membership” (2009: 1270). There are two fea-
tures of particular note about this last group. First, although generally smaller
than the other two, it is a profile that characterizes some 26 percent of younger
members surveyed in the six countries, a substantial figure in itself, and one
that rises to 30 percent in Germany and 45 percent in Hungary. Second, as

Bruter and Harrison (2009) emphasize, precisely because these data concern
younger members, it would be misleading to regard their professional ambi-
tion as something that might have developed with age or with length of
membership. Rather, it is likely to have been one of the factors that stimulated
membership to begin with. In other words, a substantial—and likely
growing—proportion of contemporary party members have joined their par-
ties with the idea of building a political career or of earning a living from or
through politics.
Because such careerist members hope to benefit in the future from the
success of the party in exactly the same way as the party in public office
does currently, it is difficult to regard such members as offering a counter-
weight to the leadership in the terms that May (1973) once hypothesized. On
the other hand, generalizing from the advice Aneurin Bevan gave to Richard
Crossman regarding advancement to the cabinet in the British Labour Party
that “there are only two ways of getting into the Cabinet. One way is to crawl
up the staircase of preferment on your belly; the other way is to kick them in
the teeth” (Crossman 1972: 32), it remains possible that some of these mem-
bers, rather than simply constituting a rather pliant organizational periphery,
will at some point attempt to exploit the more “May’s-Law-like” attitudes of
the moral-minded members to advance their own intraparty careers.8
Whether as cause or effect, the shrinkage in membership levels has been
accompanied in many cases by increased formal empowerment of those mem-
bers who remain. Parties in an increasing number of polities have begun to open
up decision-making procedures, as well as candidate- and leadership-selection
processes, to “ordinary” party members, often by means of postal ballots
(Hazan and Rahat 2010, esp. 89–105; Gallagher et al. 2011: 338–48). Rather
than witnessing the withering away of the power of the old party on the ground,
therefore, what we see instead is the apparent democratization of internal party
life, with the ordinary members beginning to win access to rights that formerly
were jealously preserved by the party elites and activists.
One of the most obvious signs of change is that “primary elections” for
party leaders have become more commonplace. Essentially unknown in 1965,
according to Pilet and Cross (2014: 457) in 2012 nearly 30 percent of party
leaders in the countries they surveyed were selected either in full member votes
or in primaries essentially open to any voter; just between 1975 and 2012, half
of the forty-four parties in their database that existed at both time points had
moved in that direction (Cross et al. 2016: 40–2). Leadership primaries have
now been used by parties in polities as diverse as Britain, Belgium, Iceland,
8
It is worth remembering that even in the 1920s, Max Weber characterized party members as
“for the most part merely [having] the function of acclamation of their leaders. Under certain
circumstances, however, they may exercise some forms of control, participate in discussion, voice
complaints, or even initiate revolutions within the party” (Parsons 1964: 408).

Ireland, France, Portugal, Italy, and Spain. In most cases, the primary elec-
tion has been confined strictly to card-carrying party members, with this right
seen as offering an incentive for party supporters to take the additional step of
joining the party. In other cases, however, the barriers to party membership
are so low as to minimize the significance of the difference between members
and casual voters, leading for example to Canadian Prime Minister Joe
Clark’s complaint that many of the participants in the 1998 Progressive
Conservative leadership selection were merely “tourists” within the party. In
the case of Labour in the UK, the primary election is open to all direct
members of the party, as well as to individual members of affiliated trade
unions who register as Labour supporters and unaffiliated individuals who
register as supporters and pay a fee, while in the case of the Democrats in
Italy, it is open to all residents, regardless of political affiliation, willing to pay
a small voting fee. In 2012 (effective for the leadership selection in 2013), the
Liberal Party of Canada moved from an all-member vote to a system in which
anyone not a member of another party and affirming that they “support the
Liberal Party of Canada” could vote without any payment at all. The selec-
tion of candidates, and the ordering of party lists, for parliamentary elections
has also often been opened up, usually by granting the ultimate power of
selection to ordinary members rather than to smaller groups of delegates.
At the same time, however, the reality of this empowerment should not be
exaggerated. Even when the final decision regarding the selection of the leader
or of parliamentary candidates is made through a membership vote, in
contrast to selection by a party convention at which numerous candidates
might be proposed over the course of the deliberation, the candidates from
among whom the members may choose in direct member votes must be
decided in advance, and commonly have been selected, or at least approved,
in ways that allow for significant input from the party in public office and its
leadership (Pilet and Cross 2014). Most directly (but also for that reason
somewhat atypically), for example, while the final choice of leader of the
British Conservative Party is made through a one-member-one-vote ballot
(OMOV), there are only two candidates, both selected by the Conservative
members of parliament. Even in the far more open Labour Party leadership-
selection process, nomination by 15 percent of the members of Labour Party
MPs and MEPs (20 percent to challenge an incumbent leader) is required in
order to appear on the membership ballot. Particularly with regard to the
selection of parliamentary candidates, member choice is frequently tempered
by the retention or introduction of a powerful veto power on the part of the
party leadership. For example, while Hazan and Rahat (2010: 136) identify
candidate selection in Canada as being “highly decentralized . . . and very
inclusive (the selectorate in the party members),” this ignores the fact that
all candidates must be approved by the central party (Canada Elections Act,
67(4)(c)).

Membership empowerment is, of course, relatively cost-free so long as
party leaders are dealing with a pliant periphery rather than an organized
internal opposition. Moreover, despite the declining organizational utility of
membership, and its substitutability by professional campaigning teams and
techniques, parties can still feel the need to foster at least some semblance of
an organized presence on the ground, not least as a result of their own
organizational legacies. Most current party organizations have not recently
been created de novo to serve the needs of their current leaders, but have been
inherited by them. Although these leaders can attempt to effect major reforms
and innovations within the organizations they inherit, as Tony Blair
attempted in the case of New Labour, there are nevertheless clear limits to
the capacity for change. If a party already enjoys a presence on the ground,
then it is unlikely that this can be easily done away with. Membership may not
be valued very highly, but a membership-oriented tradition cannot easily be
dismissed. In addition, and as part of this legacy of the past, membership may
also imbue the party leadership with a sense of legitimacy. In Sweden, for
example, “the parties seem to want to maintain the image of a mass party,
with a positive membership development being taken as proof that the party is
perceived as a viable channel for political representation” (Pierre and
Widfeldt 1994: 342), and a similar imperative clearly undergirded the major
membership drive undertaken by the British Labour Party following Blair’s
election as the new party leader.
In the case of new parties, and most especially new parties in new democ-
racies, however, it is unlikely that a party on the ground will be assiduously
cultivated (Kopecký 1995; van Biezen 1998). Indeed, even among older
democracies with strong membership traditions, new parties often play
down the need for an organized membership. The extreme case is of course
the Dutch Freedom Party (PVV), led by Geert Wilders, which had just one
member, Wilders himself, but which still managed to become the third largest
party in the Dutch Parliament in the 2010 elections (and the second largest in
2017, albeit with only about 13 percent of the vote). A bit more than a decade
before the founding of the PVV, Silvio Berlusconi’s Forza Italia began essen-
tially as a division of Berlusconi’s Fininvest corporation, with flanking
“clubs” (Club Azzurro—not unlike the Primrose League of the nineteenth-
century British Conservatives), rather than a traditional membership organ-
ization, giving rise to the idea of a “business firm party,” with “a lightweight
organisation with the sole basic function of mobilising short-term support at
election time” (Hopkin and Paolucci 1999).
Particularly with the rise of the internet, not only as a means of communi-
cation but also as a medium for commerce, many parties have begun experi-
menting with forms of “membership lite” (Kosiara-Pedersen et al. 2017)
that allow individuals to register their support at a lower cost, and with a
lower sense of commitment, than expected of full members (Scarrow 2013).

Susan Scarrow (2015: 30–1), in her analysis of party membership, lists six
distinct varieties of connection between parties and their supporters, ranging
from traditional individual members, through light members, cyber-members,
sustainers, social media followers and friends, down to the news audience
(“who may sign up to receive updates . . . such as newsletters [or] who access
the party web page . . . incur no obligations towards the party [and who may]
never communicate back to the party”). The PPDB reports the statutes of
roughly one third of the almost 100 parties for which data are reported to
recognize some form of “formal affiliation with reduced obligations and
reduced rights.” In Britain, for example, both the Conservatives and the
Liberal Democrats established supporters’ networks (Fisher et al. 2014).
Often marketed in the same way as membership in supporters’ clubs for
football teams, these allow the party to communicate directly with its
supporters—and generally also allow the supporters a privileged channel of
communication to the party and among themselves. But while furthering the
party’s capacity to use its members and supporters as a “focus group” for
trying out policy ideas, and perhaps “auditioning” candidates, like the focus
groups in the world of product marketing, they allow input and feedback
without ceding decision-making authority.
Still, for most of the long-established parties in Western Europe, the party
in public office cannot avoid the presence of a party on the ground. However
troublesome to the leadership it might prove to be, a mass membership is part
of the party tradition, and increasingly required by law as a prerequisite to
receiving the benefits of official party status.9
Given this legacy, how can the primacy of the party in public office be
successfully asserted? In some cases, there may be no problem. To the extent
that the party on the ground assumes the features of a party in public office in
waiting, there is less likely to be any conflict between the two. No longer
ideologically disparate, as originally hypothesized by May (1973), and no
longer necessarily enjoying conflicting interests and ambitions—since the
active members of the shrunken party on the ground are often aspirant leaders
or party professionals—the membership is unlikely to challenge the primacy
of the party in public office. Moreover, even if elements of the membership do
set up in opposition to the leadership in public office, their impact can often be
limited by democratizing decision making and by undermining the particular
privileges accruing to militants and delegated party bodies. OMOV decisions
rarely threaten party leaderships, and can prove even less threatening when
the party electorate is widened to include adherents and supporters. It is in this
9
The PVV, for example, has chosen to forego state subsidies rather than comply with Dutch
party law concerning its internal structure. Whether this was based on a desire to avoid the
encumbrance of a membership organization, or on a principled opposition to the state forcing
taxpayers to finance parties with which they do not agree, is of course open to interpretation.

sense that the empowerment of the party on the ground remains compatible
with, and may actually serve as a strategy for, the privileging of the party in
public office.10
A second strategy that can help ensure domination of the national party by
the national party in public office involves the promotion of a “division of
labor” between the national party in public office, on the one hand, and other
elements of the party, on the other. One possibility, following more or less
directly from the strategy of nominal empowerment of the base, has been
identified with a “franchise model” (Carty 2004) in which the party on the
ground has considerable autonomy in the selection of candidates, and perhaps
over some elements of the presentation of the national party “brand,” while
the center retains tight control over the main elements of the “brand,” in
particular national policy. This may be coupled with what Eldersveld (1964)
identified in the American case as “stratarchy,” in which different and mutu-
ally autonomous levels coexist with one another, with a minimum of authori-
tative control, whether from the bottom up or from the top down. Local or
regional parties work primarily at the local or regional level, enjoying almost
exclusive control not only over candidate and leadership selection, but over
the policies, programs, and strategies to be pursued within their own territor-
ial limits as well. The national party, on the other hand, dominated by the
party in (national) public office, is also free to develop its own policies,
programs, and strategies, unhindered by the demands and preoccupations of
the party on the ground.
Some degree of local autonomy in the selection of candidates has long been
in evidence (e.g., Bille 2001, who shows this to be true of PR as well as single-
member district systems). Early work on candidate selection found that even
the British Conservative Party, with authority highly centralized in the hands
of its leader,11 was reluctant to impose on its constituency parties, in large
measure because the party on the ground appeared to make loyalty to
the leader a key criterion for their own decisions (Ranney 1965). Even when
10
The 2015 selection of Jeremy Corbyn as leader of the British Labour Party, despite the
overwhelming opposition of the Parliamentary Labour Party, is effectively an exception that
proves the rule. Corbyn required nomination by at least thirty-five MPs, at least some of whom
apparently wanted the contest to represent “the full spectrum of voices in the party and hoped
members would dismiss the crazy views of the left in favour of a more mainstream candidate,”
ignoring the fact that, in the words of John McTernan, a former advisor to Tony Blair, “Political
parties are full of suicidally inclined activists and clearly some Labour members are suicidally
inclined” (Dathan 2015). One of those who “lent” Corbyn their vote, “former acting party leader
Margaret Beckett, [later] ruefully owned up to being a moron” (BBC UK Politics, September 12,
2015). There is very little the party in public office can do to protect itself from its own
miscalculations. For more on the Corbyn phenomenon, see Chapter 7.
11
Sometimes characterized, paraphrasing a nineteenth-century description of the Czarist
constitution, as “autocracy, tempered by assassination.”

the local party on the ground was not reliably loyal, the possibility that its
members might simply stop working could be a sufficient deterrent to central
interference, except when local internal divisions threatened to undermine the
existence of the local organization itself (e.g., McRae 1974 on Belgium). In
these cases, however, it was generally the local activists organized in party
meetings who were choosing the candidates. When the OMOV principle is
adopted at the constituency level, especially with an expansive sense of
“membership,” and coupled with a central right of veto, the nominal
empowerment of the local party on the ground is unlikely to represent a threat
to the autonomy of the central party in public office. In Ireland, for example,
the increasing use of OMOV in the selection of parliamentary candidates has
been accompanied by greater central office control of the number of candi-
dates to be nominated in each district as well as by an enhancement of the
center’s rights to add names to the lists or to veto those locally nominated
(Weeks 2008). Reform of candidate-selection procedures by the British
Labour Party at the end of the 1990s (to require selection from a centrally
developed list of approved candidates) was motivated at least in part by a
desire to weaken the position of constituency activists in favor of a more
professional approach “spearheaded from the centre” (Shaw 2001: 4). Simi-
larly, although Canadian parliamentary candidates are selected in
constituency-level all-member votes, the party leader’s approval is required
before the candidate chosen can use the party’s name on the ballot—a
requirement that has been used by the party leaders to require central pre-
screening and approval of potential candidates. Moreover, open membership
rules mean that party membership may swell in election years, effectively
drowning out the voices of long-time activists (Pruysers and Cross 2016).12
The more full-blown strategy of stratarchy is, as its American “origin”
suggests, most appropriate in federal systems, but given that even nominally
unitary constitutions usually involve elements of multi-level governance, it
need not be limited to them. In Canada, with the exception of the NDP,
federal and provincial parties are largely independent of one another
(Thorlakson 2009), and (unlike in the United States or Germany) neither
the national Liberals nor the national Conservatives are federations of their
provincial counterparts. Although there is too little evidence as yet to evaluate
the extent of this pattern systematically, a comparison of regional party
systems in Italy and Spain (Wilson 2012), and Ireland (Bolleyer 2009) suggests
that it is an increasingly common scenario. In Italy, there is evidence of
regional presidents, and to some extent directly elected mayors, using their
12
For example, membership in the Canadian Liberal Party, which in 2014 cost CAD$10, more
than quintupled between 2002 (a non-election year) and 2003 (when there was leadership
selection), falling to two-and-a-half times the 2002 level in 2004 (a parliamentary election year),
only to fall to below its 2002 level in 2005 (another non-election year) (Cross 2015: 54).

positions “to develop a personalised control over their regional party”
(Wilson 2016: 76, see also Fabbrini and Lazar 2016).
In contrast to the expectation that coalition formulas adopted for the
national government would be reflected at the regional level (Roberts 1989,
but see Ştefuriuc 2009 who shows that “congruence of party composition of
government coalitions across levels is in itself neither desirable nor undesir-
able for regional political actors”), even when not numerically necessary
(reflecting the suspicion that “inconsistency” would be electorally costly),
one possible indicator of stratarchy would be increased long-term heterogen-
eity of regional coalition formulas. In this case, rather than representing the
use of a region by the national party to “experiment” with a possible new
coalition formula, diversity would indicate increased freedom for regional
(or local) party organizations to adapt to local circumstances without refer-
ence to the national leadership, and conversely also less reason for the
national party to feel constrained by the needs of its regional affiliates.
Detterbeck and Renzsch (2003: 260 and 263), for example, find that while
only 9 percent of German Land governments (23 percent of Land coalition
governments) formed between 1969 and 1990 were incongruent with the
national coalition (including at least one party from the national coalition
and at least one party from the national opposition), the corresponding figures
for 1990–2002 were 30 percent and 49 percent. While the Detterbeck and
Renzsch finding is, as they recognize, at least partially driven by German
reunification, more recent data from Hanna Bäck and her associates tell the
same story for a wider range of countries (and for Germany after reunifica-
tion); four of seven countries show an increase in the proportion of incongru-
ent coalitions, with only the Netherlands—which started with all coalitions
being incongruent—showing a decrease (Table 3.4).
Greater autonomy of regional party organizations is not, of course, the
same as dominance of the party in public office. Each regional organization is
T A B L E 3 . 4 Incongruence of regional and national coalition
before and after 1999
Country % incongruent regional % incongruent regional

coalitions through 1999 coalitions after 1999
Austria 6.3 (16) 44.4 (18)

Belgium 50.0 (6) 100.0 (6)
Germany 31.8 (44) 48.6 (35)
Netherlands 100 (36) 83.3 (24)
Spain 15.0 (40) 22.9 (35)
Sweden 25.0 (20) 25.0 (40)
UK 50.0 (4) 50.0 (6)
Source: Data provided by Hanna Bäck and reported in Bäck et al. 2013.
Data range: Austria 1991–2009; Belgium 1995–2009; Germany 1990–2009;
Netherlands 1991–2007; Spain 1991–2007; Sweden 1998–2006; UK 1999–2007

its own system, reproducing the three faces found at the national level—albeit
generally with a weaker central office, and less physical separation between
the party in public office and the party on the ground. But, on the one hand,
the regional party in public office may reproduce at regional level the same
strategies to gain dominance over the party on the ground that are found at
the national level. Indeed, Karin Bottom (2007) suggests that national cartel-
ization makes it more likely that local parties will “engage in their own form
of cartelization.” And, on the other hand, even if the party on the ground
maintains strong control over its own party in public office, one consequence
of the stratarchic model is to insulate the national party in public office from
regional intrusion regardless of which regional face of the party is dominant.
In other words, even if the party on the ground does flourish in “franchised”
or “stratarchical” settings, in the end, it remains on the ground, being linked
to the national party in public office only through its often highly constrained
control of that party’s composition and leadership. It is to the latter that
primacy increasingly belongs.
POLITICS AS A PROFESSION
That party memberships are now smaller and less active makes it increasingly
plausible to believe that professional ambition has become a driving force
behind membership recruitment, and that to join a party with the idea of
eventually living off politics is no longer as unrealistic an ambition as it might
have been during the heyday of the mass party, when memberships were
substantially bigger and when affiliate organizations also offered a reservoir
of potential candidates or other likely party professionals.
The plausibility of aiming for a career in politics has also been enhanced by
the growth in the size of the political class and its administrative networks
(Borchert and Zeiss 2003; Sundberg 1994). This is partly due to the additional
layers of political offices that have been created over time, including new
European, regional, and local levels of government. Particularly at the
regional and local levels, not only have the numbers of positions grown, but
their pay has increased so that positions that once required their incumbents
to hold “real jobs” in order to support themselves can now support middle-
class lifestyles on their own. It is also due to the growth of full-time paid
positions both within the party central office and as aides employed by the
party in public office discussed above. Overall, more and more elected and
appointed officials can now earn a more or less comfortable living as profes-
sional politicians. It also follows from the expansion of public and private
funding for consultancy, expertise, and other forms of administrative support

that now help parties to acquire office and to manage public affairs. Some of
this growth has been referred to earlier, and it is one of the trends that have led
observers to speak of the “parliamentarization” or “governmentalization” of
parties. Already by the end of the 1980s, the numbers of staff employed by the
main parties in Western Europe had more than doubled with respect to the
1960s (Mair 1994: 6–7), a trend that has continued in subsequent years
(Scarrow et al. 2017). In Denmark, the number of staff employed by MPs
and parliamentary groups increased from 214 to 273 in the short period
between 1994 and 2000 (Jensen 2003: 88), while in the UK, the number of
staff employed by MPs increased by more than 56 percent (to almost 3,000)
between 2000 and 2010. In Switzerland the number of cantonal parties
employing full-time staff grew from twenty-three in 1960 to eighty-six in
1998 (Wiesli 2003: 380). There is little to suggest that these patterns are
exceptional.
The size of the political class and its administrative networks has also been
enlarged through the increased opportunities for patronage that have opened
up in recent years, particularly for parties in government. There are two senses
in which patronage is relevant to parties and party government (Kopecký and
Mair 2012). The traditional form is patronage as an electoral resource, with
jobs and other goods being distributed at the local level in exchange for
electoral support. At least within the advanced democracies, this type of
patronage appears to have declined in recent years, given that the types of
benefits on offer by the parties are unlikely to have the appeal they once
enjoyed in poorer and less mobile communities. This type of patronage is also
of little relevance to the nature and size of the political class.
The second form is patronage is an organizational resource. Patronage in
this sense is less a form of vote gathering and more a way of building party
organizational networks in the public and semi-public spheres, a means of
improving the steering and control capacity of parties in government, and
potentially a source of employment both for aspiring and for retiring members
of the political class. It is this type of patronage that has tended to expand in
recent years, not least as a result of the practice of allowing parties and
politicians to appoint members to the boards of the vastly proliferating
independent agencies that now play a decisive role in the making and imple-
mentation of public policy. In the Irish case, for example, it has been esti-
mated (by Clancy and Murphy 2006) that the huge growth in state agencies in
recent years has generated 5,000 new public appointments, most of which are
at the discretion of the parties in government and their ministers. Here we see
a suitable outlet for ambition.
One obvious, but for that reason often overlooked, feature of modern
democratic life is that it has long seemed both stable and enduring. For the
generation of political actors and activists that grew to political maturity in
established democracies of Western Europe or North America since the

1960s, the survival of democracy, and hence the survival of the democratic
political system with all its various structures and institutions, could more or
less be taken for granted. Even in the newer democracies of Southern and
East Central Europe, the attainment of democracy in the 1970s and 1990s
might reasonably be regarded as a political end state—something that was
finally achieved and that would become consolidated. Democracy, as Linz
and Stepan (1996) once put it, became the only game in town, whether in
Western Europe in the 1960s, in Southern Europe in the 1980s, or in East
Central Europe in the 1990s. This also meant that the institutions of dem-
ocracy had become consolidated, and this contributed to the plausibility of
building long-term careers in politics. In other words, just as young appren-
tices might see the profession of plumber or mason as offering a long-term
career, or just as students might see the profession of teacher or doctor as
offering a career that would grow and likely become more rewarding with
time, so too could young activists and ideologues look at politics—at least
within the mainstream parties—as something that could sustain them for a
lifetime, and that could eventually be highly rewarding both financially and
in terms of status.
Such perspectives were also facilitated by the improving financial and
occupational circumstances of the political class. Across the democratic
world, as Borchert and Zeiss (2003) clearly show, reforms from the 1950s
and 1960s onwards offered MPs and ministers substantially increased salaries
and social benefits. In the Finnish case, for example, in the 1950s, MPs’
salaries were linked to the income categories of civil servants and ministers,
and while this was changed in 2000, it was accompanied by a salary increase
for MPs of between 16 and 42 percent. At the same time, from 1992, MPs
could earn a full pension after just fifteen years of legislative service
(Ruostetsaari 2003: 117–19); similarly, French deputies who serve three
terms (fifteen years) are credited for their pensions as if they had served for
twice that long (http://BBC.co.uk/2/hi/uk_new3s/politics/7961849.stm). In
Norway and Sweden, where MPs also earned a generous salary and benefits,
a full pension could be acquired following twelve years in the legislature, and
in Denmark after twenty years (Narud 2003: 312; Hagevi 2003: 366–7; Jensen
2003: 99). British MPs who lose their seats or stand down at a general election
are entitled to a “resettlement allowance” worth at least 50 percent of their
annual salary (£74,000 in 2015). Similar provisions exist in most modern
parliaments, although salaries continue to vary considerably from polity to
polity, as do the benefits in kind that accrue to the MPs. Although systemat-
ically comparable data are difficult to assemble, Table 3.5 shows 1995 parlia-
mentary base salaries that were between almost four and over thirteen times
the median net family income in the same countries.
All of this serves to reduce the risks that were often associated with the
pursuit of a political career, with reasonable pensions and other severance

T A B L E 3 . 5 MP annual base salary divided by median net
household income
Country MP annual base salary/median net household

income: 1995*
Belgium 3.94
Ireland 5.87
Italy 13.22
Netherlands 5.24
UK 3.92
Note: * using an exchange rate of 1 euro = $1.34 (end of year rate)
Source: Brans 2012, table A.3; Eurostat, table EU-SILC survey [ilc_di04]
arrangements offering a particularly important compensation for loss of

office. Risks are also reduced by the efforts of parties to use other appoint-
ments to compensate for any premature loss of income by MPs. Former
parliamentarians in good standing with their party in the Netherlands, for
example, will often be offered the position of a city mayor, a relatively
lucrative appointment, or might become a royal “commissioner” in one of
the Dutch provinces. Moreover, as in the British case, opportunities in the
private sector are also often available to former MPs and ministers. According
to data gathered by Eggers and Hainmueller (2009), for example, Conservative
MPs accumulated substantially more wealth through outside employment both
during and after their terms of office than did the Conservative candidates
who failed to get elected, with incumbency more than tripling the chances that
a Conservative MP would become a director of a public company. Even more
significantly for our argument—because it runs counter to the naive expectation
that such post-political private-sector employment would be the preserve of
Conservative politicians—former Labour ministers also frequently find lucra-
tive private-sector paychecks (Guardian, May 17, 2011).
Moreover, the increasing availability of paid positions in the party
network—in party foundations, in parliament, assisting MPs or MEPs in
their offices—means that a professional career in politics can be started
quite early, and that it does not always require a sideways shift from another
sector. Although figures are scarce, and any conclusions are likely to depend
on anecdotal or ad hoc evidence, it appears that an increasing number of MPs
in the parliaments of Europe started their careers as assistants or paid advisors
to more senior MPs or ministers. In the Swedish case, for example, Hagevi
(2003: 360–1) speaks of the growing significance of pre-parliamentary experi-
ence in professional politics, as well as charting the growing proportion of
professional politicians in the Swedish Riksdag—from fewer than 15 percent
of all Riksdag members in the immediate post-war parliament to more than
30 percent by the mid-1990s. Careers are also fostered by longevity in

parliament, of course. In Britain, for example, as Jun (2003) notes, the more
recent generations of MPs have tended to enter parliament earlier and to stay
longer than their predecessors, with the median length of service of British
MPs increasing from fourteen years for those leaving the profession in the
period 1945–74 to eighteen years in the period 1974 to 1992. “Once elected,”
he concludes (Jun 2003: 174), “most MPs want to stay in Parliament for the
rest of their professional careers and give up their seats only involuntarily.”
Although the professionalization of politics may not be a new phenomenon,
in that attention to the rise of the career politician and party bureaucrat goes
back to Weber and Michels at the beginning of the twentieth century, the
evidence cited above, as well as the work of Borchert and Zeiss (2003) and Best
and Cotta (2000) more generally, indicate that both the scope and the scale of
the opportunities to live from politics have increased substantially since the
1960s. Looking only at the development of base parliamentary salaries since
1976, Table 3.6 shows not only that service in a national parliament now
provides a salary well above the national median for full-time workers, but
in addition that the provision of salaries that on their own can underwrite an
upper-middle-class life is a relatively recent phenomenon; since 1976 salaries
have been growing far more rapidly than GDP per capita, with the only
exceptions being in countries in which parliamentary salaries were already
high by the beginning of the period. Overall, the rewards in terms of political
career prospects are now greater and more widely distributed than before,
while the risks involved in choosing such a career have been sharply reduced.
T A B L E 3 . 6 Parliamentary base salary 1976, 2012
Country Parliamentary salary US$ Parliamentary base

salary/GDP per capita
1976 2012 1976 2012
Italy 2,000 182,882 0.319 9.340

Japan 22,880 268,813 4.120 6.573
New Zealand 16,500 116,945 2.524 4.633
United States 44,600 175,000 5.181 3.987
France 21,640 86,155 3.340 3.623
Germany 16,080 119,796 2.380 3.354
Ireland 6,000 120,674 1.462 3.296
UK 11,250 106,488 1.849 3.141
Canada 18,000 160,891 2.268 3.093
Australia 21,512 195,804 2.908 3.067
Sweden 17,600 103,420 2.233 2.845
Norway 18,212 122,250 2.915 2.519
Switzerland 3,300 82,736 0.323 2.033
Spain 3,000 43,948 0.590 1.578
Source: Herman and Mendel (1976); Independent Parliamentary Standards Authority (2012)

Two important implications follow from these developments. First, and
most obviously, the growth in political careerism will inevitably have made a
difference in how political leaders and aspirant leaders perceive politics and
the role of party therein—and here we can refer back to the Bruter and
Harrison (2009) findings noted above. Second, and perhaps more import-
antly, it will also have made a difference in how the leaders and aspirant
leaders of any one political party perceive their counterparts in the other
parties. To repeat von Beyme’s (1996: 149) observation quoted in Chapter 1,
“the new political class as a transfer class [is] privileged in two respects: by
being the only elite sector which determines its own income, and by organiz-
ing state-support for the organizations which carried them to power, e.g., the
parties.” Put differently, the growth in opportunities for exercising the pro-
fession of politics has come about as a result of policies and decisions imple-
mented by the professional politicians themselves. It is parliaments, and the
parties and MPs within these parliaments, that have legislated for their own
salaries, benefits, and pensions. It is also the parliaments that have legislated
for the introduction of state subventions to parties, and that have allocated the
resources that have allowed the parties in public office, as well as their
organizations outside parliament, to increase their staffing levels and to
engage consultants and other professionals. These benefits have not come to
parties exogenously, except in the sense that they are eventually inherited by
later generations of party leaders; rather, they are privileges that have been
wrought by those for whom they are intended.
This also means that they have been devised cooperatively rather than
competitively. Although fringe parties and populist parties may often rail
against the privileges of the political class, even at the same time as they
enjoy their benefits, there has probably been no case in modern democratic
history in which the benefits—salaries, pensions, resources, etc.—accruing to
parties and their leaders in public office have been pushed through against the
will of one or more mainstream parties. When proposals for party benefits risk
being seen in partisan terms, they tend to be withdrawn, and it is only when
there is all-party agreement that such changes are implemented. Especially
given that salary increases for politicians have become increasingly unpopular
with voters, it is only through cooperation (or collusion) that the now more
common practice of disguising increases in compensation by packaging them
as “allowances” or “expenses” can succeed.13 Within the mainstream at least,
13
In many cases, the fear of popular backlash has meant that nominal salary increases have
only allowed MPs to maintain roughly constant real salaries in the face of inflation. On the other
hand, to take the British example, while the basic nominal salaries of British MPs increased by
roughly 1240 percent between January 1972 and January 2007, the maximum payable allowance
increased by a whopping 11,690 percent (House of Commons Library briefing paper SN/PC/
05075).

the provision of public aid for parties and their leaders is, perforce, a largely
depoliticized domain. To paraphrase van Biezen (2004), the stated goal is
always to protect parties in general as public utilities, rather than parties in
particular as partisan enterprises.
This in turn helps to frame relations between parties and between their
leaders, which is why the growth in the opportunities provided for politics as a
profession will have made a difference in how the leaders of any one party
perceive their counterparts in the others. The notion that “there is less differ-
ence between two deputies, one of whom is a revolutionary and the other is
not, than between two revolutionaries, one of whom is a deputy and the other
is not,”14 attributed originally to Robert de Jouvenel (1914: 17), clearly
applies with particular vigor to contemporary political leaderships, in that
there is substantially less difference between two holders of public office, one
of whom is, for example, a Christian Democrat, than between two Christian
Democrats, one of whom is a holder of public office. Over and above the
necessity for cooperation and collusion that is often fostered in many polities
by the dispersal of institutional power and by the need for informal as well as
formal coalitions, political leaders and MPs, by virtue of their institutional
roles, also share common needs and interests, and these include facilitating the
pursuit of their profession and the work of their organizations. In this sense,
the primacy of the party in public office in any one party is made easier by the
primacy of the parties in public office in other parties. Political leaders and
aspirant leaders may not share their parties in common, but they do share
their profession.
CONCLUSION
It is important to understand that parties now tend to be dominated by a body

of leaders who derive their living and status from politics, and who therefore
will be inclined to adapt their organizations and their politics to the need to
enhance such a living. As Theda Skocpol (1992: 40) put it in another context,
In 2009, members of the French National Assembly received a comparatively low base salary of
€5180 per month. But they also received an expense allowance of €5790 per month for lodging,
travel, and entertainment—plus free first-class rail travel within France and forty return flights per
year between Paris and their constituencies.
14
Il y a moins de différence entre deux députés dont l’un est révolutionnaire et l’autre ne l’est
pas, qu’entre deux révolutionnaires, dont l’un est député et l’autre ne l’est pas.

politicians and administrators must be taken seriously. Not merely agents
of other social interests, they are actors in their own right, enabled and
constrained by the political organizations within which they operate . . .
Both appointed and elected officials have ideas and organizational and
career interests of their own, and they devise and work for policies that
will further those ideas and interests, or at least not harm them.
In so doing, party leaders, and the party in public office more generally, now
play the strongest role in defining the character of party organizations and the
parameters of party competition. Parties dominated by such leaders are clearly
far removed from both the image and day-to-day practice of the mass parties
that preceded them. Indeed, in contrast to the mass parties, as well as in
contrast to the more transitional catch-all parties that flourished in the
1960s, 1970s, and into the 1980s, contemporary parties have become less
organizations of and for political activists and members who hire professionals
to help them, and more organizations of professional politicians with an
activist corps of supporters who aspire to succeed them, and with a broader
army of cheerleaders who help carry them to office. Today, for example, anyone
with an interest in how party organizations function might be less advised to take
an interest in the activist layers and in the role and activities of the party
conference—themes that were still strongly favored in research projects during
the late 1970s (Reif et al. 1980; Van Schuur 1984; Minkin 1978)—and encour-
aged instead, along with Poguntke and Webb (2005), Webb et al. (2002) Hazan
and Rahat (2010), and Plasser and Plasser (2002), to take a closer look at the
candidates and the leaders, at how they are selected, and at how they build and
dispose of their staffs and professional advisors.
The argument presented here also reflects the conclusions drawn by Cotta
and Best (2000: 520–5) regarding the transformation of legislative recruitment
over the past century and a half. Developing a taxonomy based on the level of
democratization, on the one hand, and the level of professionalization in
legislative recruitment, on the other, they draw a valuable broad-brush dis-
tinction among four phases of democratic development (see Figure 3.1).
Democratization
Professionalization Low High
Low The dignitary The
functionary
The free political The professional
High entrepreneur politician
F I G U R E 3 . 1 Cotta and Best’s typology of legislators


The first phase corresponds to the period in which restrictive electoral laws
and a narrowly defined franchise were combined with popular deference to
and reliance on the recruitment of privileged notables—or dignitaries—into
parliament. Parties, when they existed, were of the elite or cadre type. The
second phase belongs to the free political entrepreneur, and corresponds to the
transition from the cadre to the mass party, with the power base of politicians
resting mainly on their ability “to mobilize the support of informal caucuses
and networks of those who control the selection process of candidates and are
able to trade votes for political influence” (Cotta and Best 2000: 525). Func-
tionaries, in turn, typify the mass party age, with paid positions in the party
apparatus itself or in one of the many affiliated cleavage organizations pre-
ceding and also facilitating (in the absence of professional MP salaries)
recruitment to parliament. As Cotta and Best suggest, these functionaries
also tended to be very loyal agents of their party principals, not least because
their election and living were directly dependent on these principals. The final
phase in their taxonomy comes, as we also highlight here, with the rise of the
professional politician—“politicians purs et durs, for whom the rules of the
political field determine their actions and expectations” (2000: 524). However,
it is not only legislative recruitment that is changed in this way, so too are the
parties, with the dominance of professional politicians within each of the main
competing protagonists effecting a transformation in the ways in which the
parties organize, as well as in the ways in which they relate to one another.
Parties and One Another
In the classic model of democratic party government (Rose 1974; Katz 1987;
see also Figures 1.1–1.3), political parties were understood to be competitors
in what was effectively a fixed-sum game that determined both the personnel
of government, and the direction that policy would take. While alternative
elaborations of this model may have differed as to whether the immediate
focus of competition ought to be the specific policy programs that the parties
put forth at a particular time, the more general ideologies they espouse, the
particular social segments to whose interests they cater, or the particular
leaders heading the party team, it was taken as necessary and proper that
the stakes of party competition were high, both for the society in terms of
public policy and for the party organizations themselves. It was the personal
and organizational costs of losing that were presumed to keep parties and
politicians “honest” and to encourage them to be responsive to their electors
and to the public at large.
From the rise of mass suffrage (and before, but that is less relevant here)
through the immediate post-war period, political competition was largely
about solidarity and identity. The principal issues were understood in terms
of conflict between well-defined groups, and the primary electoral strategy
was one of mobilization of each party’s natural clientele, accomplished not
only by stressing within-group solidarity, but also between-group hostility.
The rhetoric was not always as harsh as Aneurin Bevan’s famous “vermin
speech”: “No amount of cajolery, and no attempts at ethical or social seduc-
tion, can eradicate from my heart a deep burning hatred for the Tory Party.
So far as I am concerned they are lower than vermin.”1 Nonetheless, the
theme of fundamental class conflict was at the core of social democratic party
programs. Moreover, this sense of fundamental conflict was not limited to the
social class divide. In a 1954 bishop’s letter, Dutch Catholics were enjoined
against “membership of the Socialist labor union, regular attendance at
Socialist meetings, regular attention to Socialist newspapers and radio broad-
cast,” on pain of being denied the sacraments (Lijphart 1968: 36). Far from
being competitors in a friendly, if sometimes intense, game, parties were more
1
Speech on July 3, 1948 at the Bellevue Hotel, Manchester.

like combatants in a war—able to form alliances of convenience if, for
example, required for coalition formation, but still fundamentally isolated
and mutually antagonistic. In the terms of Lipset and Rokkan’s well-known
adaptation of Talcott Parsons, the oppositions that define party differences
were predominantly at the l (local regional) and especially at the i (ideological)
ends of the g-l and a-i axes (Lipset and Rokkan 1967: 10). Even if the
underlying cleavages generally were no longer matters over which people
killed one another, as they clearly had been when they were formed in earlier
centuries, they still were not divides that could easily be bridged.
Experience in wartime grand coalitions or of being together in exile or in
the same concentration camps had, by the 1950s, already softened some of the
mutual hostility and suspicion of opposing parties. During the forty years
after Lipset and Rokkan published their “frozen cleavages” hypothesis, how-
ever, the pattern of mutual hostility and isolation across cleavage lines
decayed even more, albeit in fits and starts in some places. These changes
took several interrelated forms. One was a shift in the nature of issues from the
i end of Lipset and Rokkan’s a-i axis, which is characterized by questions of
ideology that are in principle non-compromisable except through mutual
acceptance of peaceful coexistence, toward the a (interest specific) end, at
which compromise in the form of splitting the difference is more realistic.2 As
binary (two-position, essentially friend versus foe) issues have been replaced
by or transformed into issue continua, a second change has been a tendency
for the policy positions of the major parties to converge; moreover, this
convergence has not always been toward the position of the median voter,
suggesting that this convergence is not just the natural outcome of rational
vote-maximizing strategies.3 Third, as substantive policy positions have con-
verged, the emphasis in electoral competition has shifted away from policy
differences (what the different parties would do) and towards differences in
personalities (sometimes identified as the “personalization,” or at the top as
the “presidentialization,” of politics), experience, or “managerial competence”
(how effective the different parties might be).
2
Religion is a classic example of an i-end cleavage; substantive compromise between
Catholicism and Calvinism, or between either of those and secularism, could only be achieved
by effectively destroying both ideologies. On the other hand, the intense political conflict between
advocates of secular schools and advocates of religious schools could be defused by the Dutch
School Law of 1920 with mutual acceptance of the principle that Catholic, Calvinist, and secular
schools all would be accorded equal rights, and afforded equal state funding. Even more, the
question of how much support to give education could easily be resolved by compromise at any
point along a monetary continuum.
3
That convergence is not simply the result of rational vote maximization is further suggested
by the fact that the theorem showing convergence to the median to be a Nash equilibrium depends
crucially on there being exactly two parties, a condition rarely met, or even approximated in
polities with proportional electoral systems.
Parties and One Another 83

These changes have, in turn, facilitated changes in the relationship among
parties. Being less strongly divided by questions of policy, parties have become
more promiscuous in their choice of coalition partners. (For example, see
Pajala 2013 on Finland.) Cooperation across the government-opposition divide
(what in two-party systems would be identified as bipartisanship) has come to
be valued as a positive trait. While electoral competition may remain intense,
and even apparently bitter at times, it has become more like the competition of
sports rivals, interested in besting but not in destroying one another, rather than
like the competition of rival armies intent on killing their enemies.
This coalition promiscuity is not unlimited, however. As we argue later, the
more frequent crossing of cleavage boundaries to find coalition partners as well
as the policy convergence that is the subject of the next section, have contrib-
uted to the rise of a variety of racist, extreme, or, as they now are commonly
characterized, “populist” parties that are treated as non-coalitionable—subject
to the kind of cordons sanitaires applied (often under pressure from the United
States) to West European communist parties in the late 1940s and 1950s.
POLICY CONVERGENCE
In the immediate pre- and post-war periods, the issues over which parties
competed tended to be based on deeply embedded social cleavages, which in
turn supported a political strategy of mobilization of a well-defined classe
gardée. This cleavage-based politics was manifested in a number of ways. One
was the degree to which lines of political division had become entrenched—the
so-called “frozen cleavages” observation of Lipset and Rokkan that “the party
systems of the 1960s reflect, with few but significant exceptions, the cleavage
structures of the 1920s” (1967: 50, italics in original), notwithstanding the
upheaval of World War II. Similarly, Galli and Prandi (1970: 18–19) observed
that the territorial distribution of strength of Italian parties in 1946 virtually
mirrored their strength in 1919, that is before the rise of the fascist regime.
Another manifestation of this political style was the sociological distinct-
iveness of the various parties’ electorates. Because of its clear connection to
the left-right axis of political contention, research focused most intensely on
social class, with the Alford Index (the percentage of the working class voting
for a party of the left minus the percentage of the middle class voting for a
party of the left) the most common metric. According to figures computed by
Dalton (2008: 148), the Alford Index for the United States in 1948 (although
declining rapidly thereafter) was over 40; for the UK it was even higher than
that in 1966; in the high 30s in Germany in the late 1950s; and in the mid-20s
in France in the mid-1950s. Data published by Oscarsson and Holmberg

(2010: 18) shows an Alford Index value of 53 for Sweden in 1960. In this
context, it made perfect sense for Peter Pulzer (1975: 102) to declare that
“Class is the basis of British party politics: all else is embellishment and
detail,” or for Lipset (1960: ch. 8) to identify elections as “the expression of
the democratic class struggle.”
Class was not the only cleavage structuring party politics, however. Reli-
gion (either Protestant versus Catholic or Catholic versus secular) was also
an important line of division. Although Dalton (2008: 159) shows that the
Alford Index equivalents for religious denomination were already somewhat
lower than those for class in the 1950s and 1960s, in the earlier days of the
twentieth century religion was in some countries a considerably stronger
determinant of party preference than was class. In England, for example, it
was reasonable to characterize the Church of England as “the Tory party at
prayer” (attributed to preacher and suffragette Agnes Maude Royden by
Bell (2014) citing O’Donnell (1990: 60), but widely quoted thereafter). Simi-
larly, until they merged into a single Christian Democratic Party, the Dutch
Catholic People’s Party (KVP) and the Protestant Christian Historical Union
and Antirevolutionary Party were largely defined by their confessional nature,
as the Staatkundig Gereformeerde Partij is still defined by its orthodox
Calvinism. Adopting the approach of looking for social cohesiveness within
individual parties,4 rather than party differences between social groups, in the
mid-1960s, Rose and Urwin (1969) found 75 percent of the seventy-six parties
in seventeen countries for which they had data to be cohesive on at least one
dimension, with religion actually a slightly more common basis for social
cohesion than class—and with language, region, and ethnicity also the bases
of party cohesiveness in some cases.
Not surprisingly, this was also manifested in the rhetoric of the parties.
These passages from the 1950 Manifesto of the British Labour Party (itali-
cized in the original) provide an example:
Our appeal is to all those useful men and women who actively contribute to the
work of the nation. We appeal to manual workers—skilled, semi-skilled and so-
called unskilled; farmers and agricultural workers; active and able managers
and administrators in industry and the public services; professional workers,
technicians and scientists; and housewives and women workers of all kinds.
The fundamental question for the men and women of the United Kingdom to
determine when they vote is this: Shall we continue along the road of
ordered progress which the people deliberately chose in 1945, or shall
reaction, the protectors of privilege and the apostles of scarcity economics
be once more placed in the seats of power, to take us back to the bleak years
of poverty and unemployment? Those years must never return.
4
The difference in approaches is illustrated by race in the United States. Although, with
roughly 90 percent support for the Democrats among African Americans, the Alford Index
equivalent for race would have been at least in the 40s, the Democrats would not be counted as
cohesive with regard to race since a substantial majority of Democrats were, in fact, white.

T A B L E 4 . 1 Percentages of quasi sentences coded in “Labour Group: Positive” for social
democratic party manifestos and “Labour Group: Negative” + “Middle Class and Professional
Groups: Positive” for liberal and conservative party manifestos, 1950–70 and 1991–2005
Percentage of quasi sentences Percentage of quasi sentences in

in “Labour Group: Positive” “Labour Group: Negative” +
category for social democratic “Middle Class and Professional
party + US Dem; Italian PCI; Groups: Positive” categories for
Canadian Lib manifestos liberal and conservative party
manifestos
Country 1950–70 1991–2005 1950–70 1991–2005

max mean max mean max mean max mean
Denmark 20.5 3.5 8.0 3.5 15.3 1.9 6.4 0.7
Finland 25.4 8.6 8.5 8.3 70.3 7.7 0.0 0.0
Iceland 13.1 6.5 0.0 0.0 11.7 2.4 2.8 0.8
Belgium 10.3 6.0 3.0 1.0 12.7 7.7 1.2 0.4
Netherlands 2.0 1.1 5.2 3.6 2.0 1.0 1.2 0.6
Luxembourg 14.7 6.1 3.9 2.3 38.3 16.3 0.6 0.4
France 6.2 4.3 4.9 2.4 6.6 2.7 0.0 0.0.
Italy 16.3 9.6 7.5 2.8 7.0 2.0 1.1 0.7
Austria 5.2 1.8 7.5 6.3
Switzerland 12.1 3.3 9.0 3.9 2.2 0.6 0.6 0.4
Great Britain 5.2 2.3 2.9 2.1 6.7 2.8 1.5 0.4
Ireland 17.2 5.2 2.2 2.1 0.0 0.0 2.3 0.6
US 7.1 4.5 3.8 2.7 3.0 2.2 0.2 0.1
Canada 8.4 1.8 9.2 3.1 2.6 0.4 1.2 0.3
Australia 23.2 14.5 6.2 2.8 3.9 0.7 1.9 0.9
New Zealand 4.0 2.5 9.7 4.4 3.3 1.4 0.6 0.1
Source: Volkens et al. (2012)
More systematically, the Comparative Manifestos Project has coded the

“quasi-sentences” in party manifestos, yielding a comparable measure of the
relative salience accorded to fifty-six categories of concerns, among which are
categories referring to social groups. The first pair of columns of Table 4.1
shows the average and maximum percentages of the codeable quasi-sentences
that fell into the category “Labour Groups: Positive” in the manifestos of
social democratic parties (plus additional parties when the largest left party,
such as the Italian Communists, the American Democrats, or the Canadian
Liberals, was not classified in the social democratic family) between 1950 and
1970 for the sixteen countries for which this calculation could be made.5 The
third pair of columns repeats this exercise for the sum of “Labour Groups:
Negative” and “Middle Class and Professional Groups: Positive,” this time for
5
Norway and Sweden are excluded because the uncodeable proportion of quasi-sentences is
missing.

T A B L E 4 . 2 Electoral volatility, 1945–65 and 1970–2004
Country 1945–65 1970–2004
Austria 5.2 15.8

Belgium 9.4 18.3
Denmark 8.7 21.1
Finland 5.0 17.5
France 16.3 21.3
Germany 12.4 16.2
Ireland 10.7 18.7
Italy 12.7 24.9
Netherlands 5.2 19.7
Norway 4.8
Sweden 5.0 20.1
Switzerland 3.3
UK 4.6 17.6
Source: Bartolini and Mair (1990) and Caramani (2006)
parties in the liberal and conservative families.6 Particularly given that these can
be relatively policy-free categories in what are customarily expected to be policy
documents, Table 4.1 shows simple group appeals to be significant—and some-
times quite remarkably significant—components of party manifestos in the early
years of the post-war era.
By the mid-1970s, this pattern had clearly begun to break down. One
unmistakable sign was the 1973 “landslide” election (Jordskredsvalget) in
Denmark, in which the number of parties represented in the Folketing essen-
tially doubled.7 More systematically, comparison of overall electoral volatil-
ity between 1945 and 1965 computed by Bartolini and Mair (1990) with
analogous figures for the 1970s through 2004 computed by Caramani (2006)
shows very substantial increases (Table 4.2).
Other changes include a dramatic decline in class voting with, for example,
the US Alford Index dropping from 41 in 1948 to 1 in 2000 and 2 in 2004 and
the British index dropping to under 15, with the figures for France and
Germany both under 10. In 1991 (and again in 2006), the Swedish figure
was 25—still relatively high, but less than half of what it had been in 1960.
While the decline in the corresponding figures for denominational voting has
not been as dramatic, they too show substantial declines. Not surprisingly,
this weakening of the relationship between social cleavages and party choice
6
Although the Christian Democratic family is often included in the “right” or “bourgeois”
bloc, it is excluded here because of the cross-class nature of Christian democracy.
7
The raw count of the number of parties represented in the Folketing went from five in 1971 to
eleven in 1973, while the “effective number of parliamentary parties” went from 3.9 to 6.9.

has been reflected in the content of party manifestos. The second and fourth
pairs of columns in Table 4.1 repeat the measures in the first and third pairs,
but this time for 1991–2005. Although there are exceptions, the general trend
(more than two thirds of the comparisons) is for references to the traditional
clientele groups of parties to have become less prominent in their manifestos.
Moreover, most of the contrary cases are the result of a single, highly excep-
tional (outlier) manifesto.
Accompanying the decline in the connection between party and location
with regard to the class and religious cleavages (in addition to the data above,
see esp. Dogan 2001), and the increase in electoral volatility, the post-war
period has also seen a constriction of the portion of the left-right policy
dimension occupied by what might be described as “mainstream” parties.
This is illustrated in Table 4.3, which compares the distance on the left-right
axis between the leading left and right protagonists as perceived by country
experts surveyed in 1984, 1995, and 2002. Although the trend has not
always been monotonic, with only five exceptions—one of which (Italy) is
not strictly comparable because of the collapse of the Christian Democrats
and their replacement as the dominant party of the right by Berlusconi’s Forza
Italia—the experts see the two major protagonists to be more similar in
left-right placement by the end of this period than they were in 1983. More-
over, four of the five remaining exceptions were among the six countries that
began the period with the most constricted left/right dimensions, and so had
little space for further contraction.
Using the measure proposed by Laver and Budge (1992) to derive left-right
party positions from the manifesto data shows much the same thing over a
longer period. Looking at all of the parties in the social democratic, liberal,
conservative, and Christian democratic families, Table 4.4 shows the distance
on the 200 point scale8 between the placement of the left-most left party and
the right-most right party within this group, averaged over elections in the
periods 1950–70 and 1996–2005. In all but five cases, there has been a decline
in the left-right range of these parties—and three of those five experienced a
major change in their party systems in the mid-1990s that resulted in the entry
into the mainstream of a new party that scored far to the right of the parties
found in the earlier period.9
8
Note that this 200 point range would only be realized if parties not only were entirely
consistent in their left or right orientation, but also failed to devote any manifesto space to
concerns, such as social justice or governmental and administrative efficiency, that are not part
of the left-right index.
9
The “standard” rational choice explanation for this convergence is convergence toward the
median voter. If this were so, the expectation would be for the left-most party to move to the right
and the right-most party to move to the left, i.e., in opposite directions. In eleven of the eighteen
cases in Table 4.4, however, the parties moved in the same direction. The same is true in ten of the
fourteen cases in Table 4.3.
T A B L E 4 . 3 Expert survey left-right party placements
Country Left-most mainstream party Right-most mainstream party Range
1983 1993 2002 1983 1993 2002 1983 1993 2002
Australia Labor 3.1 Labor 4.2 Labor 4.7 Country 7.8 Country 8.3 Liberal 7.8 4.7 4.1 3.1
Austria SPÖ 3.0 SPÖ 4.2 SPÖ 4.1 ÖVP 5.8 ÖVP 5.8 ÖVP 7.0 2.8 1.6 2.9
Belgium PSB 2.5 PSB 3.3 PS 1.8 PVV 7.8 VLD 7.1 5.3 5.3
Canada Liberal 5.3 Liberal 4.6 Liberal 5.3 PC 6.5 PC 7.0 Reform 8.9 PC 6.6 Alliance 8.7 1.2 2.4 4.3 1.3 5.4
Denmark SD 3.8 SD 3.6 SD 3.5 KF 7.3 KF 7.3 KF 7.5 3.5 3.7 4.0
Finland SDP 3.0 SDP 3.8 SDP 3.9 KOK 7.2 KOK 7.1 KOK 7.7 4.2 3.3 3.8
France PS 2.6 PS 3.5 PS 3.2 RPR 8.2 RPR 7.6 RPR 7.0 5.6 4.1 3.8
Germany SPD 3.3 SPD 3.1 SPD 3.9 CSU 7.9 CSU 7.0 CDU/CSU 6.6 4.6 3.9 2.7
Great Britain Lab 2.3 Lab 3.8 Lib Dem 4.7 Lab 5.2 Lib Dem 3.6 Con 7.8 Con 7.5 Con 8.1 5.4 4.7 2.8 2.9 4.5
Ireland FF 6.3 FF 5.3 FG 6.2 FG 6.8 FG 6.7 FF 6.5 0.5 1.4 0.2*
Italy PCI 1.6 PDS 1.7 DS 2.6 DC 5.4 DC 5.9 FI 7.7 3.8 4.2 5.1
Netherlands PvdA 2.6 PvdA 3.6 PvdA 4.0 VVD 7.4 VVD 6.9 VVD 8.1 4.8 3.3 4.1
New Zealand Labour 3.8 Labour 5.3 Labour 4.3 National 6.0 National 7.0 National 7.2 3.2 2.7 2.9
Norway DNA 3.0 DNA 3.5 DNA 3.6 H 7.7 Fp 9.4 H 7.8 Fp 9.1 H 8.2 Fp 7.8 3.7 6.4 4.3 5.6 4.6 4.4
Sweden SD 2.9 SD 3.4 SD 3.8 M 7.7 M 8.1 M 8.5 4.8 4.7 4.7
USA Dem 4.8 Dem 3.5 Dem 3.2 Rep 6.8 Rep 6.5 Rep 8.2 2.0 3.0 5.0
Note: * in a sense, even this overstates the range, given that the left and right parties actually switch positions
Source: 1983: Castles and Mair (1984); 1993: Huber and Inglehart (1995); 2002: Benoit and Laver (2006)

T A B L E 4 . 4 Mean left/right positions of the left-most party in the social democratic
family and the right-most party in the liberal, conservative, and Christian democratic
families, 1950–70 and 1996–2005
1950–70 1996–2005
Social Dem Right Range Social Dem Right Range
Sweden 44.1 40.4 84.5 10.9 37.7 47.6

Norway 32.8 7.1 39.9 29.9 9.0 38.9
Denmark 17.8 40.8 58.6 19.1 30.5 49.6
Finland 24.0 41.9 65.9 13.1 19.6 32.7
Iceland 18.9 30.5 49.4 20.6 16.9 37.5
Belgium 21.3 19.8 41.1 22.2 3.7 25.9
Netherlands 25.4 19.4 44.8 14.9 19.5 34.4
Luxembourg 32.0 0.4 32.4 23.0 6.0 29.0
France 34.3 21.7 56.0 14.7 5.2 19.9
Italy 28.7 5.8 34.5 2.7 39.0 36.3
Austria 22.2 18.3 40.5 18.5 7.3 25.8
Switzerland 23.2 29.8 53.0 35.8 51.7 87.5
Great Britain 27.1 2.3 29.4 2.9 18.4 21.3
Ireland 26.4 33.6 60.0 14.1 6.7 20.8
US 17.0 5.4 22.4 4.6 27.7 32.3
Canada 38.9 1.9 40.8 28.1 21.9 50.0
Australia 19.3 23.8 43.1 4.2 36.0 40.2
New Zealand 34.6 6.3 40.9 27.3 33.6 60.9
The two overall trends that these data illustrate—a weakening of the social
bases of party politics and a weakening of ideological distinctiveness of parties—
had many roots. One clearly was change in the structure of societies themselves.
There has been a significant shrinking of the industrial working class, the
traditional home of class-based politics. For example, the French industrial
working class, which represented 51 percent of the electorate in 1951, was only
30 percent in 1988 (Dogan 2001: 101). Similarly, the proportion of German
workers in manual occupations dropped from 51 percent in 1950 to barely 40
percent in 1985 (Dalton 1988: 82). The flip-side of this has been the growth of a
large self-identified middle class. In 1963, Butler and Stokes (1969: 67) found 32
percent of the British to identify themselves as “upper-middle,” “middle,” or
“lower-middle” class; in the 1995 wave of the World Values Survey in Britain,
68.8 percent identified themselves as “upper-middle” or “lower-middle” class. In
1971, 46.2 percent of respondents to the 1971 Dutch Parliamentary Election
Study identified themselves as “middle class” or “upper-middle class”; in 2003,
the figure had risen to 79.2 percent. (Although it fell back to 60.9 percent in 2006,
this was still substantially above the 1971 figure.)10 Similarly, 34.8 percent of
10
Dutch Parliamentary Election Study Cumulative Dataset, 1971–2006 (ICPSR 28221),
Principal Investigator(s): Aarts, Kees, University of Twente (Netherlands), and Dutch Electoral

respondents to the 1957 Norwegian election study identified themselves as
middle class, a proportion that had grown to 56.1 percent in the 2001 study.11
Another trend has been an increase in social mobility. Using a four-class
ranking of occupations, Noble shows that ten-year career mobility for British
men between 1971 and 1981 (45 percent changing category) was more than
twice what it had been for 1953–63 (21.6 percent) (Noble 2000: 38). Looking
at a series of studies over the period from 1949 to 1984, Noble (2000: 40) also
reports a steady increase in upward intergenerational mobility across the
manual-non-manual divide (and a decline in downward intergenerational
mobility, contributing to the shrinking of the manual working class overall).
There also has been a weakening of ties to organized religions. To take the
case that was regarded as the archetype of social segmentation along confes-
sional lines, between 1967 and 1986 the share of the Dutch population
identifying with the Dutch Reformed Church was cut in half, from 30 percent
to 15 percent; by 2006, it had been halved again, to 7.3 percent. While the
proportion identifying as Roman Catholic only declined from 37 percent to 31
percent between 1967 and 1986, by 2006 it had fallen to 24 percent; moreover,
the percentage of Catholics attending church as least once a week plunged
from 75 to 26 in 1986 and 21 percent in 2006. More simply, between 1971 and
2006, the proportion of Dutch survey respondents classified as “no religion”
grew from 29.4 percent to 54.3 percent. Similarly, between 1989 and 2009,
attendance at mass by Catholics in Germany was cut in half, from 28 percent
to less than 14 percent (St. Leger 2009); between 1967 and 1992 the proportion
of Germans of any Christian denomination saying that they attend church
“every or almost every Sunday” fell from 25 percent to 10 percent (Dogan
2002: 143).
According to Putnam (2002: 408–9) “between the 1970s and the late 1990s,
church attendance declined in virtually every European country.” The World
Values Survey shows that the proportion of the population attending religious
services once a month or more fell between 1981 and 2000 by 15 percent in
Australia, 11 percent in Canada, 6 percent in France, 18 percent in Ireland, 15
percent in the Netherlands, 18 percent in Spain; in Denmark, Iceland, and
Sweden, the figure was already below 15 percent in 1981 (Halman et al. 2008:
210). Even where there is some evidence of increased religious belief, it
Research Foundation; Todosijevic, Bojan, University of Twente (Netherlands); van der Kaap,
Harry, University of Twente (Netherlands) . Data provided by the Inter-university Consortium
for Political and Social Research.
11
Rokkan, Stein, and Henry Valen. Norwegian Election Study, 1957. ICPSR07288-v2. Ann
Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2010-11-
30; Sapiro, Virginia, and W. Philips Shively. Comparative Study of Electoral Systems, 2001–2006.
ICPSR03808-v2. Ann Arbor, MI: Inter-university Consortium for Political and Social Research
[distributor], 2008-07-01.

appears to be “not in the official, traditional, institutional way, but in a
personal, unofficial way” (Halman and Riis 2003: 10). Thus even if religious
beliefs still contribute to party choice, the institutional reinforcement that
characterized earlier decades is no longer there.
Some of this social change, and the lessened ideological distinctiveness of
parties, followed from the acceptance of the welfare state, its success, and then
the social and fiscal consequences of that success. First, with the acceptance by
the “bourgeois” parties of the right and center that the state had either a moral
responsibility or a prudential incentive to assure a social safety net, an
important bone of contention was removed from the political agenda; while
the parties continued to disagree over the appropriate levels and conditions of
provision, the binary ideological question was answered, and replaced by a
continuum along which compromise and convergence were possible. Second,
public subsidy or direct provision of such services as old age pensions,
unemployment insurance, secondary and tertiary education, and medical
care dramatically reduced the differences in life experiences and life chances
between social groups, as well as increasing the possibility and likelihood of
inter- or even intragenerational movement across what had previously been
much more impenetrable class boundaries. Particularly before the advent of
cable and direct satellite broadcasting, so that most people could only receive
a quite limited number of national channels, the growth of television as the
dominant source of information and entertainment for the vast majority of
people produced a great commonality of (often vicarious) experience that cut
across cleavage divisions.
Another root was the growing hegemony of neoliberalism among profes-
sional economists, and then even more among economics reporters and politi-
cians (see Blyth 2002). This hegemony was arguably even greater than the
position previously held by Keynesianism, but even more significantly here,
whereas Keynesianism called for an active governmental role in managing
the economy, neoliberalism called for a withdrawal of government from eco-
nomic management, and hence effectively for the removal (or at least the de-
emphasis) of issues like employment and inflation from the party political
agenda.
Both the social changes and this ideational shift are reflected in the mani-
festo data. Table 4.5 shows the average share of manifestos devoted to con-
cerns in the economic cluster for the parties on which Table 4.4 was based.
Overall, the average proportion of social democratic manifestos devoted to
economics declined by 4.9 percent (just over 20 percent of its original value);
while the decline in “bourgeois” manifestos was nearly as large (4.0 percent,
or 15 percent of its original value).
Policy convergence has also been driven by practical concerns. The welfare
state proved to be very expensive. In part this was the result of shifting
expenditures that had previously been in the private sector into the public

T A B L E 4 . 5 Proportions of manifestos devoted to the economic cluster
1950–70 1996–2005
Social democratic 24.2 19.3

Liberal, conservative, Christian democratic 26.2 22.2
sector, thus increasing the latter substantially. While this shift might be con-
tested endlessly on ideological grounds, in aggregate economic terms it makes
little difference whether doctors or pensions or teachers are paid by the state or
by private parties, unless the amount they are paid changes—and conservative
arguments to the contrary notwithstanding, there is sound evidence that at
least for some social programs, public provision is significantly more efficient
(more of the desired outcome relative to expenditure) than is private provision.
The economic consequences of the welfare state went well beyond this simple
shift from one ledger to another, however. First, welfare states not only
changed the way in which services were paid for, they provided more service
in absolute terms; perhaps the main justification for the welfare state, after all,
was to provide for the welfare of those who would otherwise be priced out of
the private market. More children stayed in school for longer; more people
received medical care (which became increasingly expensive as a result of
technological advances, often themselves made possible by state support of
research); the elderly lived longer and had more adequate retirement incomes.
Further, that children were in school for longer meant that they entered the
workforce later, while declining birth rates (themselves in part an indirect
effect of social policies including both the welfare state and the push for gender
equality) meant that there were fewer young people entering the workforce at
all. At the same time, the success of the welfare state in providing health care,
as well as other contributors to longevity including environmental controls and
workplace health and safety regulations, meant that there were an increasing
number of people who survived their working years and expected to be
supported in their retirement. The inevitable result was a smaller proportion
of the population in work and supporting a larger proportion of the population
(both the young and the old), a trend mitigated only by the entry of more
women into the workforce and into the ranks of taxpayers and by the arrival of
working-aged immigrants who entered the workforce and the tax base.
Particularly in the immediate post-war (and early welfare state) years, rapid
economic growth and “baby-boom” birth rates made pay-as-you-go and
deficit finance of welfare state programs seem possible. When economic
growth slowed to sustainable rates, however, and the demographic changes
just described began to take hold, the deficits and unfunded liabilities of the
welfare state contributed to severe debt crises in some countries—which

threatened to convert progressive taxation from redistribution of income from
the rich to the poor into redistribution of income from all productive elements
of society to the (often foreign) bond holders—and a fear that such crises
might be just around the corner in many others. Simply, it was no longer
possible for the left to promise more services without raising taxes, or for the
right to promise reduced taxes without loss of services.
Institutional changes also contributed to a constriction of the policy space
in which parties could compete. This was especially true within the EU. The
Single European Act, the Maastricht Treaty, and its successors, the Economic
and Monetary Union culminating in the adoption of the euro and the creation
of a European Central Bank, moved control over a wide range of economic
policies to Brussels and Frankfurt. Similarly, the Schengen system moved
much of immigration policy to Brussels. The enhanced scope and assertive-
ness of the European Court of Justice moved yet other questions out of the
control of national governments. Both within and outside of the EU, the move
toward independent central banks again furthered this trend, as did the advent
of the World Trade Organization.
All of these institutional changes had the effect of taking areas of policy out of
the realm of party contestation. This has also increasingly been done directly,
through the growth of “free” or “conscience” votes in otherwise highly discip-
lined parliaments. Taking the UK as an example, data clearly show that both the
number and range of subjects decided by such votes have grown significantly.
Without addressing the larger question of the democratic legitimacy of such
votes, they clearly represent another example of what S. E. Finer (1975: 18)
(referring to referenda) characterized as the “Pontius Pilate of British politics”—
allowing the government, and it might be added the opposition, to “wash its
hands” of an issue by claiming that it is not a party issue at all.
Many of these developments were epitomized by the 1999 joint statement
by British Prime Minister Tony Blair and German Chancellor Gerhard
Schroeder, “The Third Way/Die Neue Mitte.” Although ostensibly a celebra-
tion of the success of social democratic parties, much of it—“companies must
have room for manoeuvre to take advantage of improved economic condi-
tions and seize new opportunities: they must not be gagged by rules and
regulations”; “corporate tax cuts raise profitability and strengthen the incen-
tives to invest . . . It helps create a virtuous circle of growth”; “the taxation of
companies should be simplified and corporation tax rates cut”; “social con-
science cannot be measured by the level of public expenditure. The real test for
society is how effectively this expenditure is used”; “the responsibility of the
individual to his or her family, neighborhood and society cannot be offloaded
onto the state”; “within the public sector bureaucracy at all levels must be
reduced, performance targets and objectives formulated, the quality of public
services rigorously monitored”—could as easily have come from parties of the
so-called right. From the other side, the Swedish Moderates (the mainstream

party of the right) essentially campaigned (successfully) in 2010 on the
platform that they would defend the social democratic welfare state more
effectively than the Social Democrats themselves (Steinmo 2010: 80).
CHANGING MODES OF COMPETITION
That party positions have converged, and that important issues have been
removed from the party political agenda, does not mean that parties have
ceased to compete, but it does contribute to a shift in the mode and basis of
competition. One example of this lies in the 2010 Swedish election to which
reference was just made; rather than appealing for support on the basis of the
policies they would change if elected, the Moderates based their appeal on the
claim that they would do the same thing, only do it better.
Returning to the manifesto dataset, there are three categories of quasi-
sentences that are particularly policy-free: “governmental and administrative
efficiency: positive”; “political corruption: negative”; and “political authority:
positive.” Table 4.6 shows the average of the total proportion of the manifesto
quasi-sentences that fell into these categories in 1950–70 and 1996–2005 for
the same mainstream parties on which the earlier tables were based. In
fourteen of the eighteen cases, this proportion has gone up, in some cases
quite dramatically; averaged over all the manifestos on which this table is
based, the proportion of quasi-sentences in these categories nearly doubled.
Along with the increased focus on such criteria as efficiency, honesty, and
competence, there has been an increase in personalization—in the case of
parliamentary systems, sometimes identified as “presidentialization.” One
indicator of this, as well as being a factor that reinforces it, has been a
dramatic shift in the way in which cabinets are described in the media.
A series of full-text searches of the Times (London) illustrate this quite clearly.
In 1949, the Times was more than sixteen times as likely to refer to the
“Labour government” rather than to the “Attlee government” or to “Attlee’s
government.” In 1952, it was more than nine times as likely to refer to the
“Conservative government” rather than to the “Churchill government” or
“Churchill’s government.” By 1996, this had been reversed, with the “Major
government” or “Major’s government” used more than seven times as often
as “Conservative government,” and 2004 when references to the “Blair gov-
ernment” or “Blair’s government” were nearly five times more common than
references to the “Labour government” (Table 4.7). While the cause of this
shift may lie in part in the spillover effects (to the press) of the exigencies of
television reporting and campaigning, its effect is to highlight the personal
qualities of the leader, rather than the policies or ideology of the party—and

T A B L E 4 . 6 “Governmental and administrative efficiency: positive,” “political corruption:
negative,” and “political authority: positive”
Social democrat Bourgeois
1950–70 1996–2005 1950–70 1996–2005
Sweden 0.0 0.2 1.0 1.4

Norway 1.7 3.5 1.1 3.4
Denmark 0.8 4.5 2.0 6.1
Finland 17.2 11.8 14.4 2.6
Iceland 10.8 6.9 14.7 8.0
Belgium 4.0 9.2 4.7 12.2
Netherlands 1.9 8.3 2.3 10.3
Luxembourg 1.6 3.7 2.4 6.8
France 3.8 3.2 8.5 4.6
Italy 3.6 21.7 6.8 19.5
Austria 3.3 9.1 0.9 4.5
Switzerland 3.6 3.7 7.3 12.5
Great Britain 1.2 14.3 4.0 14.9
Ireland 8.1 7.3 19.1 7.8
US 6.2 12.8 7.0 12.5
Canada 3.5 3.8 3.8 12.2
Australia 5.3 22.9 8.1 17.9
New Zealand 1.2 5.9 2.8 10.3
T A B L E 4 . 7 References to (party) government and (prime minister) government in

the Times (London) 1949, 1952, 1996, 2004
Prime minister (Party) government (PM or PM’s) government
1949 Attlee 94.3% (133) 5.7% (8)

1952 Churchill 90.1% (145) 9.9% (16)
1996 Major 11.3% (138) 88.7% (1083)
2004 Blair 17.5% (219) 82.5% (1030)
Source: 1949 and 1952 from the Times digital archive, Infotrac; 1996 and 2004 from http://infoweb.newsbank.com
as well to suggest a kind of “Führerprinzip” in the making of party decisions.

Moreover, it may have the effect of casting interparty deals in the guise of
personal accommodation, which is generally socially approved, rather than
compromise of principles, which finds much less social support.
An additional change in the media reporting of politics in general, and of
election campaigns in particular, has been the increased emphasis on the
horserace, or more generally the “game,” quality of the campaign (who is
leading and by how much; who has gained or lost ground since the last report
or the last poll; what strategies are the contestants pursuing) rather than the

substance of political differences (Kaid and Strömbäck 2008: 430). In part this
is simply the result of data availability: with improvements in polling tech-
nologies, it became more possible to track short-term changes in party stand-
ings, and once a poll was commissioned, simple business interests dictated
that it be reported. Moreover, a new poll is by definition new.
It is not clear whether this focus on the horserace has come at the expense of
substantive coverage of policy differences or “objective” candidate qualifica-
tions, or as an addition to it. Blumler and Coleman (2010: 143), write about
“an ever-increasing emphasis on politics as a game” as a characteristic of
British political communication. Brants and van Praag (2006), report that the
horserace content of campaign coverage on the Dutch program NOS-journaal
increased from practically nothing in 1956 (for lack of the necessary technology)
to 18 percent in 1986 and 43 percent in 2003, while the proportion of coverage
they identified as “substantive” stayed relatively constant over time. Norris
(1997: 75) identifies movement “from cleavage-based and issue-based
conflict towards character-based ‘personalization’ of politics” and “from a
campaign revolving around party platforms towards a poll-driven focus on
the ‘horse-race’ strategy” as two of the key changes in post-1945 British
political communication.
Horseraces are, of course, examples of real competition, as are contests
between sports teams. Nonetheless, the casting of the competition between
parties in these terms corresponds to a significant change in thinking about
that competition away from that suggested by the principal-agent models of
Figures 1.1–1.3. For one thing, it downplays policy differences; the question is
not which party has the better policies, but which party is more skillful in
promoting its policies. Moreover, like reports of athletic contests (or other
entertainment activities), it tends to emphasize “stars,” and thus again devalues
enduring party differences over policy, ideology, or clientele, in favor of atten-
tion to naturally transitory questions of the personalities at the top of the parties.
To continue the sporting analogy, not only the leading players/candidates but
also the coaches, who in the political case are the media consultants and
advertising professionals, assume greater importance. But like coaches, who
may well be more attached to the game than to any particular team, and who
may change teams over time, these campaign professionals tend to promote a
similarity among party strategies—and those strategies are likely to entail the
downgrading of principle-driven policy in favor of marketable images.12 In this
respect, it is unclear whether the move of Saatchi & Saatchi from Conservative
to Labour referred to in Chapter 3 should be understood as cause or effect of the
hollowing out of interparty differences.
12
Of course this is not an entirely new phenomenon. Quintus Tullius Cicero, for example,
advised his older brother, the more famous Marcus Tullius Cicero, to avoid taking stands on
issues for fear of making enemies (2012 [64 BC]).

COALITIONS
There has been a widely observed trend for elections to take on a bipolar
character (one identifiable coalition opposing another), even in countries with
proportional representation and multi-party systems. What has been less
widely observed is a marked increase in the fluidity of national coalition
formulas, including the installation of governing coalitions that would in the
earlier decades of the post-war era have been regarded as unthinkable, and a
similar increase in the proportion of regional coalitions that are inconsistent
with the coalition formed at the national level. The divergence of national and
regional coalitions has already been discussed (see Chapter 3). In this section
we consider the fluidity of coalition formulas at the national level.
If considered in enough detail, virtually every government is unique, both in
the sense that it (the party or parties included) occupies a different proportion
of the parliamentary seats and in the sense that different people occupy
ministerial posts. Coalitions will also differ with respect to the allocation of
ministries among the cooperating parties, and especially may differ with
regard to the party of the prime minister. Here, however, we define a govern-
ing formula simply as the list of parties forming the government, and a new
governing formula to have been created whenever a coalition is formed that
represents a combination of parties that have not previously (since the count
began, in our case in 1947) governed together. Thus the replacement of an
ÖVP-SPÖ coalition (Klaus I) by an ÖVP single party government (Klaus II)
in Austria in 1966 counts as a new formula, as does the replacement of that
government by an SPÖ government (Kreisky I) in 1970, while the replacement
of the Kreisky IV SPÖ government by another ÖVP-SPÖ coalition (Sino-
watz) in 1983 does not, even though the party occupying the post of federal
chancellor was different than in the earlier ÖVP-SPÖ coalitions. Similarly, the
initial addition or subtraction of a small party may represent a new governing
formula, even if the core of the coalition remains unchanged.
Table 4.8 summarizes the number of new government formulas in thirteen
long-standing European parliamentary democracies, comparing the twenty-year
periods 1960–79 and 1996–2015 as well as showing data from the period of
1947–59, when parliamentary government was re-establishing itself after the
war. While only a few cases suggest an acceleration of the rate at which new
formulas are tried, neither do these data suggest that the process of innovation
has stopped. And of course one must remember that once a formula has been
tried, perhaps even in the 1940s, it can never again be counted as “new.” As an
alternative way of viewing this process, Table 4.9 shows the total number of
different (not necessarily new) governing formulas employed in the same coun-
tries in the periods of 1947–69 and 1993–2015. (By this method, the same formula
can be counted in both periods.) Although the differences are small, the pattern is
quite consistent; there are more formulas used in the later period.

T A B L E 4 . 8 Numbers of new formulas, 1947–59, 1960–79, 1996–2015
1947–59 1960–79 1996–2015
Austria 1 2 2
Belgium 3 9 4
Denmark 3 4 1
Finland 10 6 5
Germany 5 3 1
Iceland 5 2 2
Ireland 3 1 2
Italy 6 6 8
Luxembourg 2 1 1
Netherlands 4 5 4
Norway 1 2 3
Sweden 2 3 2
UK 2 0 1
T A B L E 4 . 9 Numbers of government formulas, 1947–69 and 1993–2015
Jan 1947–Dec 1969 Jan 1993–Dec 2015
Austria 2 3
Belgium 5 7
Denmark 5 7
Finland 15* 8
Germany 8** 3
Iceland 5 4
Ireland 3 5
Italy 8 10
Luxembourg 2 3
Netherlands 6 7
Norway 3 4
Sweden 2 3
UK 2 3
Note: * includes six different formulas by 1951, two of which never appear again;
** includes four different formulas by 1957, none of which appear again
Perhaps most significantly, coalitions spanning what were once thought to

be unbridgeable cleavages, in particular those dividing liberals and Christians,
and especially those dividing social democrats from both of these blocs, have
become far more common. This process had already taken place before World
War II in the classic consociational democracies, with a Catholic-Liberal-
Social Democratic coalition in Belgium as early as 1916; in the Netherlands,
this happened in the 1930s, with a Catholic-Protestant coalition in office since
1918 expanded in 1933 to include Liberals, and then expanded again in 1939
to include Social Democrats; in Switzerland, the Social Democrats entered the
Federal Council in 1943, and after absence from 1953 to 1959, the four major

parties (Social Democrats, Christian Democrats, Free Democrats-Liberals,
and Swiss People’s Party) have governed together with only the exception of a
few months in 2008—occasioned by a schism within the Swiss People’s Party.
Although the Christian and Social Democratic parties in Austria were briefly
in coalition immediately after the establishment of the republic in 1918, the
real change from opposition to cooperation came after World War II, with
coalitions of the ÖVP and SPÖ in power from the re-establishment of parlia-
mentary government in 1945 through to April 1966, from 1978 to 2000, and
again from 2007 to 2017.
Following its abandonment of Marxist theories in 1959, the German Social
Democrats entered a coalition government with the Christian Democrats in
1966 (lasting until 1969), and then again participated in Großen Koalitionen
with the CDU/CSU from 2005–9 and from 2013–17, transforming what may
once have appeared to be an exception into a normal governing formula. In
Italy, the wall between Christians and socialists was breached in 1963 with the
entry of the PSI into the first Moro government,13 and although it did not
come to full fruition, the period of the compromesso storico in the late 1970s
eroded the antipathy between the Christian Democrats and the Communist
Party. In Norway, the division between socialist and bourgeois blocs was
breached in 2005 when the Center Party joined a coalition led by the Labor
Party. From the foundation of the Fianna Fáil party in 1926 until 1989, the
Irish party system was well described as “Fianna Fáil versus the rest,” and
indeed Fianna Fáil had often expressed a preference to be in opposition rather
than to form a coalition government with any other party. Against this
history, the 1989 decision of Fianna Fáil to form a coalition with the break-
away (from Fianna Fáil and Fine Gael, founded in 1985) Progressive Demo-
crats represented a significant shift in Irish politics, but even more the 1992
coalition of Fianna Fáil and Labour (a core member of the traditional “rest”),
more than simply being one more new coalition formula, was indicative of a
major expansion of the range of coalition possibilities.
Many of these openings were responses to external threats (or numerical
necessity), and may have been seen as temporary expedients. Once a precedent
was set, however, it remained a possibility for the future. On the one hand, the
more realistic the possibility of entering government, the greater incentive for
party leaders to act in ways that would not repel potential coalition partners,
and to attempt to structure their internal party arrangements in ways that
would not hamper this effort. On the other hand, the more parties behave in
this way—regarding any existing coalition formula as provisional and any new
coalition formula as possible—the more reasonable it becomes for citizens to
13
Although the PSDI (Italian Social Democratic Party) entered government in 1954, it, like
the Portuguese Social Democratic Party, is actually a liberal/conservative party.

think of government by “the parties” rather than government by any particu-
lar, and transient, coalition of parties—and for parties that are excluded from
these arrangements, or that exclude themselves, similarly to attack “the par-
ties” rather than the particular coalition in power at the moment. The sense
among the leaders of one mainstream party that they may in the near future be
coalition partners with the leaders of virtually any other mainstream party,
and the perception among the leaders of all of the mainstream parties that they
are jointly under attack from what we have called the “anti-party-system-
parties,” then furthers their sense of solidarity and their sense of separation
from the rest of society. It may also incline them to “circle the wagons,” and to
withdraw collectively into the warm embrace of the state.
As we suggested in Chapter 3, the rise of politics as a profession, and the
recognition that they have common career needs and incentives, has led mem-
bers, and particularly leaders, of parties in public office to become a political
class, not just in the sense of a descriptive category of people working in the
same “industry,” but in the far stronger sense of a group who recognize that
they share an interest that is in competition with the interests of other groups,
and that this common interest transcends the competition that continues to
exist among members of that political class. Politicians within a single party
continue to compete among themselves for advancement within that party, and
the members of the parties in public office of the various parties continue to
compete with one another for the (shrinking) advantages of being in office
rather than in opposition, but the politicians of all parties also cooperate to
protect and advance their common interests as professional politicians.
That argument was largely focused on politicians as (self-interested) indi-
viduals, and underscored the continuing validity of Jouvenel’s observation
about the similarity of “two deputies, one of whom is a revolutionary and the
other is not.” In this chapter, however, we have suggested that the parties
themselves have become more similar in their policies, and more open to
cooperation with one another. It is not just the relations among politicians as
individuals but among the mainstream parties as collectivities that have
changed, in both cases in ways that tend to mitigate or constrain the conflicts
within the political class and among the parties, but that also have the
potential to widen the gap between the political class and the mainstream
parties on the one hand, and much of the rest of society on the other, and
moreover to alter the relations between the parties and the state.
Parties and the State
For most of their history, political parties were understood in political theory,
by political scientists, and in public or constitutional law to be external to the
state. Particularly in pre-liberal states, parties were often, and accurately, seen
as subversive of the political order as institutionalized in the then existing
constitutional arrangements. In early liberal regimes, parties were originally
seen as sinister combinations or factions, that is as groups or organizations of
citizens intent on pursuing their private advantage at the expense of their
fellow citizens and of the common interest, which it was the job of the state to
protect. With the increasing importance of representative assemblies, on the
one hand, and the expansion of suffrage, on the other, the legitimacy and
necessity of the kind of organization that political parties embody both in
parliament and in the electorate came to be widely accepted. But even if
parties were no longer seen to be antithetical to the existing state, they were
still understood to be separate from it.
At least into the second half of the twentieth century, parties were generally
seen to be organizations firmly rooted in civil society: articulating and aggre-
gating demands from society; channeling political participation by citizens;
recruiting and certifying contenders for representative offices. With the advent
of proportional representation (PR), parties of necessity were recognized in
law, but often only as organizations of candidates. Even after “party govern-
ment” was recognized as the dominant “legitimizing myth” of democratic
governments (Castles and Wildenmann 1986), and cabinets came to be occu-
pied primarily, if not exclusively, by individuals who owed their ministerial
appointments to their positions as party leaders, one still talked about parties
“occupying government offices” and “taking control of the state” without
actually becoming part of it. In some countries, this separation of party and
government was institutionalized in provisions asserting that members of
parliament should be answerable only to their individual consciences or to
the nation as a whole, rather than to their parties,1 and in some countries
1
For example, article 38 of the Basic Law of the German Federal Republic: “Members of the
German Bundestag . . . shall be representatives of the whole people, not bound by orders or
instructions, and responsible only to their conscience”; (Die Abgeordneten des Deutschen
Bundestages werden in allgemeiner, unmittelbarer, freier, gleicher und geheimer Wahl gewählt.

perhaps even more in a requirement that ministers vacate their parliamentary
seats, either for the duration of the parliament or at least while they held
executive office.2
Starting after World War II in Europe (somewhat earlier in the United
States), but particularly starting in the last quarter of the twentieth century,
there has been an accelerating trend to redefine the relationships between
parties and civil society on the one hand, and between parties and the state, on
the other. Focusing on the relationship between parties and the state, this
redefinition can be seen from two perspectives. From the point of view of the
state, parties have become more intrusive—they have “entered” the state—in
the sense that they and their leadership cadres have taken on an increasing
range of official roles and responsibilities. Rather than simply taking temporary
command of the state apparatus (as exemplified by the stylized, mid-
twentieth-century view of the British executive—each ministry a unified hier-
archy of permanent civil servants headed by a permanent secretary, above
whom there would be a single political minister, who not only would be the
only person to be replaced with a change of government, but indeed would
be the only person with explicit connection to any party3), individuals for
whom a connection to party politics is a prerequisite rather than an impedi-
ment to office have become far more prominent in the policy-making and
policy-implementing apparatus of the state.
If the state has been increasingly entered by the parties, however, the parties
have also become increasingly entered by the state. The growing recognition
of the centrality of parties in modern democracies has been accompanied by
increasing legal recognition and privileges, but also by increasing regulation.
Quite aside from the question of whether some parties might be banned on the
basis of their “subversive ideologies” (recognizing that in the longer run of
history, the novelty is not the banning of subversive parties, but rather the
acceptance that parties are not all subversive by their very nature as parties),
in many countries not only the financial arrangements and the internal
organizational/decision-making structures of parties, but even acceptable
choices of their candidates for office, have been prescribed by law, and
according to norms that treat parties as semi-public or even fully public
institutions, rather than private associations. Parties have been explicitly
assigned a range of public functions that would in earlier times clearly have
Sie sind Vertreter des ganzen Volkes, an Aufträge und Weisungen nicht gebunden und nur ihrem
Gewissen unterworfen), or article 27 of the Constitution of the Fifth French Republic: “Any
specific instruction to a member of Parliament [from an outside body] is null and void” (Tout
mandat impératif est nul. Le droit de vote des membres du Parlement est personnel).
2
Such a requirement can also be seen as enforcing a separation of executive and legislative
branches of government.
3
This, of course, ignores the cadre of junior ministers and parliamentary private secretaries.
Parties and the State 103

been within the purview of the state. Moreover, governments have accepted a
growing list of treaty obligations that effectively treat parties as if they were
part of the state structure. In other words, it is not so much that the parties
have come to excessively penetrate the state, or that the state has come to
excessively control the parties, as it is that the two have come to interpenetrate
to such a degree that the boundary between them has become unclear.
This is not entirely a new phenomenon. The Jacksonian use of public jobs
to reward party supporters (classic party patronage) can be seen as an early
American antecedent of the blurring of the boundary between state and party,
as more recently can the practices of Proporz in Austria or lottizzazione in
Italy. As early as the 1950s, Otto Kirchheimer had observed what he identified
as a cartel of centrist catch-all parties directed at maintaining their privileged
access to public office, as well as a detachment of parties from their social
foundations and a tendency for politics to be reduced to “state manage-
ment.”4 Nonetheless, in the period since the 1970s, the pace of change
accelerated, and the magnitude of change accumulated, to the point that it
is better characterized as leading to a fundamentally different set of relation-
ships, and it is only after these new relationships are recognized that hindsight
allows the significance of early antecedents to become clear.
PARTY LAWS
All democracies have laws that bear on political parties. On one hand,
although there is some dispute as to the long-term direction of causality (do
electoral laws condition the kinds of parties and party systems that develop
(Duverger 1986), or do the existing parties enact electoral laws that are
conducive to their own interests (Colomer 2005)?), there is general agreement
that, in the short term, electoral laws have a significant effect on parties. On
the other hand, for parties to be able to do business in the modern world, they,
like all associations or corporations, must have legal personality in order to
have bank accounts, contract for services, own or rent premises, and so forth.
As is true for other juridical persons as well, the laws that make this possible
will not only enable, but also constrain. “Party laws” or the “constitutiona-
lization of political parties,” however, go beyond this to treat political parties
4
In contrast to our model, however, Kirchheimer saw the model of “government by party
cartel” as driven either by the main parties’ fear of each other (e.g., Austria) or by their need to
collaborate in the face of a threat from anti-system (opposition of principle) parties (e.g., France
and Italy). While not denying the power of these external threats, our argument is driven more by
shared interests than by shared threats.

as a distinct category of organization, subject to more or different regulation
than other entities, notwithstanding that those other entities may from time to
time also be active in the political sphere.
While guarantees of freedom of association are common in democratic
constitutions, and while parties certainly have benefitted from these provi-
sions essentially since the establishment of liberal (often pre-democratic)
government, the explicit recognition of political parties in constitutions is
relatively recent. As van Biezen (2009) points out, for example, the constitu-
tions of the Weimar Republic and of pre-war Austria include mention of
parties, but primarily with reference to assuring the political neutrality of civil
servants and judges. Other early constitutional references to parties were
essentially byproducts of constitutionalizing PR as the country’s electoral
system (e.g., article 31 of the 1944 constitution of Iceland), although the
Finnish example shows that it is possible to implement PR without legally
recognizing the existence of durable organizations that nominate or support
candidate lists or ever using the phrase “political party” (Törnudd 1968: 36).5
Modern party laws—that is, “laws specifically designed to regulate the life of
party organizations” (Müller and Sieberer 2005: 435)—were widely enacted in
the United States during the Progressive era (roughly 1890–1930), although
when compiled they were generally classified under the heading of “electoral
law.” While laws regulating various aspects of political activity—for example,
the British Parliamentary Elections Act of 1868 (31&32 Vict. 125) or Corrupt
and Illegal Practices Prevention Act of 1883 (46&47 Vict. 51)—were enacted in
many European countries in the nineteenth century, extensive party laws were
only adopted in a few countries in the 1960s and 1970s (Germany: 1967;
Finland: 1969; Austria: 1975; Spain: 1978 (Avnon 1995: 287); as well as
Portugal: 1977) and again in the years following the fall of the Berlin Wall
(Bulgaria: 1990; Czech Republic: 1991; Estonia: 1994; Hungary: 1989; Lithu-
ania: 1990; Macedonia, Poland: 1990; Romania: 1996; Russia: 2001; but also
Israel: 1992 and the UK: 2000 (Karvonen 2007: 451–3)).
The enactment of these laws reflected recognition of a tension in the
democratic experience. On the one hand, it was increasingly recognized that
parties had become essential to the operation of democratic government.
Among the functions of parties in modern electoral and representative dem-
ocracies are the recruitment, selection, and presentation of candidates; even if
independent candidacies are allowed, as a practical matter a party nomination
is virtually a prerequisite for election.6 In most countries, parties dominate
5
The Finnish Parliament Act of 1906 refers only to “voters’ associations” (at least fifty
enfranchised citizens supporting a list of not more than three candidates in a particular
constituency) and “electoral alliances” of such associations.
6
For example, of the 648 members of the British House of Commons in February 2017, only
four identified themselves as “independent”—and of these, only one (Sylvia, Lady Hermon

political campaigns, defining the issues (both which issues will be prominent
and what positions will be presented to the voters), providing most of the
actual propaganda, and receiving the lion’s share of media attention. Between
elections, parties play central roles in the organization of government, and
provide important venues for popular discussion of political issues and the
formation of public opinion. Even if free and fair competition among parties
is not by itself the definition of democracy (e.g., Schumpeter 1962), parties are
sufficiently central to a democratic constitutional order that, for example, the
German Constitutional Court has recognized them as “institutions of consti-
tutional law” (verfassungsrechtliche Institutionen). The implication is that the
continued vitality of parties must be guaranteed if the continued vitality of
democracy is to be assured.
On the other hand, however, parties can also pose a threat to democracy.
Both the fascist regimes that preceded the enactment of significant party laws
in Austria, Germany, Italy, and Spain and the communist regimes that
preceded the enactment of significant party laws in the countries of Eastern
and Central Europe were founded on parties that had overthrown democratic
governments or had collaborated with foreign occupations. In other coun-
tries, such as the United States, party laws were introduced in response to
widespread corruption involving the parties.7 As Avnon (1995: 296) put it,
party laws were intended “to ensure, on the one hand, free association in
political parties as a basic tenet of a democratic regime and, on the other, to
ensure the state’s ability to ban parties whose activities represent a threat to
the democratic nature of the state.”
At the risk of some oversimplification, the content of party laws can be
classified under three main headings, although in practice there is more
interconnection and overlap than this simple classification might suggest.
The first category concerns the regulation of parties as organizations, and
addresses questions of membership and internal structure and decision mak-
ing. The second concerns the regulation of parties as political actors based in
civil society and as contestants in elections, and addresses questions of the
allowable content of party programs and campaign practices, but also ques-
tions of the rights, privileges, or obligations accorded to parties that are
denied to (or not required of) individual citizens or other organizations. The
third category concerns the activities of parties in government, and addresses
questions of patronage and other possible uses (or abuses) of state resources
representing the Northern Ireland constituency of North Down) had been elected as an
independent. (Of the other three, one was elected for Labour and the other two for the SNP.)
At the same time, of the more than 600 members of the Italian Chamber of Deputies, eleven
identified themselves as “non iscritti ad alcuna componente,” but all of them had been elected
from party lists.
7
See, for example Riordon (1995 [1905]) or Orth (1919).

for partisan advantage, requirements for the formation of party groups in
parliament (and the advantages that accrue to them), and restrictions on party
switching by MPs during a parliamentary term.
While the particular content of these laws can have a significant impact on a
party, four additional points are of special significance here. First, as laws of
general application, they impose a similarity of circumstances—incentives
and constraints—on all parties within a given system. Second, notwithstand-
ing their apparent neutrality—and in some cases, without even the pretense of
neutrality—these laws may in practice work to secure the position of one
group of parties against possible challenges by parties outside of that group.8
Third, in their justification and content, even those laws that formally treat
parties as private associations, increasingly appear to treat them in reality as
public rather than private bodies. And fourth, although they may be con-
strained by constitutional provisions and constitutional courts, these laws,
while imposed on the parties, are also, like all laws in modern democracies,
written by the parties.
PARTIES AS ORGANIZATIONS
With regard to parties as organizations, legal regulations appear to be evolv-

ing through three successive models of party organization. The earliest is
rooted in the model of the cadre party. While pre-dating most party laws,
this understanding continues to have a residual effect. In this model, parties
are primarily organizations of candidates for office and those who already
hold office, attempting to coordinate their activity and to maximize their
influence (Schumpeter 1962; Schlesinger 1994; Aldrich 1995, 2011), or per-
haps a bit more broadly as organizations of candidates and/or office holders
8
Two example from the United States can illustrate this. At the national level, the Federal
Election Campaign Act provides far more generous public support for “major parties” (defined as
those that won at least 25 percent of the vote at the previous presidential election—and this
effectively means only the Democrats and Republicans) than is given to “minor” or new parties.
As a state example, the election law of Delaware (15 Del 4502) specifically assigns (by name) the
first two positions on the ballot to the candidates of the Democratic and Republican parties.
Similarly, Article 127 of the Spanish Ley Orgánica 5/1985, de 19 de junio, del Régimen Electoral
General, Texto Consolidado, Última modificación: 1 de noviembre de 2016 provides for the state to
subsidize party electoral expenses, while Article 127 bis calls for the parties to receive advances
equal to 30 percent of the subsidy they received in the previous election—but of course this means
that while parties that were strong in the previous election are entitled to large advances, weaker
parties get less, and new parties get nothing at all.

plus their supporters. This understanding of party remains prominent in the
academic literature particularly in the guise of the “electoral professional
party” (Panebianco 1988) or the “modern cadre party” (Koole 1994), as
well as in many rational choice analyses of parties, even as it is recognized
that the cadre party per se is obsolete as an organizational form. Because it
minimizes the importance of formal party organization, as opposed to infor-
mal cooperation, it effectively excludes state regulation of the internal
decision-making procedures of the party, except perhaps within the standing
orders of a parliamentary chamber. Indeed, this model was characterized by a
reluctance even to recognize the existence of parties for fear of being drawn
into internal conflicts, although this would not necessarily prevent a party
from adopting rules and having them become legally enforceable in the same
way as the rules of any other private association might be interpreted as a
legally enforceable contract among the members. While those rules might
include a requirement of some form of member participation in decision
making, however, they might equally permit, as John Ramsden metaphorically
described the traditional (pre-Margaret Thatcher) constitution of the British
Conservative Party (analogizing to Czarist Russia), a system of “autocracy
tempered by assassination.”
The second model, derived from the mass party of integration, has been the
most prominent philosophical basis for party laws in Europe. In this view,
parties are associations of citizens who work together on a long-term basis to
advance their collective interests and to secure the election of their preferred
candidates. In structural terms, this reverses the dominant/subordinate roles
of party members and candidates/officials in the cadre model. From this
perspective, party organizations are important media through which citizens
exercise their rights of speech, petition, and political organization. It is, for
example, in this context that the German Basic Law (Art. 21) assigns to
parties the duty to “participate in the formation of the political will of the
people.”9 Thus, if the state’s obligation to secure the political rights of citizens
extends beyond merely refraining from itself interfering with their exercise, to
include the obligation to protect citizens against private impairment of their
rights, then the state may be justified in regulating parties, and in particular in
requiring some level of control by the parties’ members.
In both of these views, parties are still seen as essentially private entities,
although in the second view their centrality to democratic government and
their public functions within that framework justify more extensive state
protection, and more extensive state regulation, than would be compatible
with liberal political rights if imposed on, for example, advocacy groups. In
9
Similarly, art. 4 of the French Constitution: “Political parties and groups shall contribute to
the exercise of suffrage.”

the third view, best typified by the United States at the state level (the level at
which most American regulation of parties takes place), but also increasingly
relevant for Europe, parties are seen as “public utilities” (Epstein 1986; van
Biezen 2004) or even as semi-public entities. While they may be recognized as
having some of the rights of independent organizations, the emphasis shifts
decisively to the range of public functions that they are expected to perform.
Rather than being seen simply as the contestants in elections, their organiza-
tions are recognized as part of the structure of elections.10 In this case, parties
are likely to be subject to even more detailed regulation—for example, rather
than merely being required to be internally democratic, their entire structure
may be prescribed in detail.11 Rather than merely being prohibited from
discriminating against ethnic minorities in their membership requirements,
parties may be left with no discretion at all concerning their membership.12,13
In one sense, these more stringent regulations can simply be seen as more
extreme or detailed regulations to facilitate the exercise of individual political
rights. In another sense, however, particularly in European states (both the
members of the EU and the members of the larger Council of Europe), the
justification for these regulations has shifted from protection of individual
10
Perhaps the clearest example of this shift is in the American jurisprudence on the subject of
the “white primary”—primary elections from which non-white citizens were excluded by party
rules. In the case of Grovey v Townsend (295 U.S. 45, 1935), the US Supreme Court ruled that the
Texas state Democratic Party was a “voluntary political association,” and therefore not
constrained by the 14th and 15th Amendments to the Constitution. Nine years later, in Smith v
Allwright (321 U.S. 649, 1944), the Court ruled that Texas’ “statutory system for the selection of
party nominees for inclusion on the general election ballot makes the party . . . an agency of the
State” (p. 663, italics added) and therefore bound by those amendments
11
To cite a relatively extreme example, Vermont state law requires that the base unit of a political
party be the town committee, elected by a town caucus that must be organized in each odd-numbered
year, and in which all “voters of the party residing in town” may participate (17 V.S.A. }}2301–20).
The law specifies that the town committee is to elect five officers (chair, vice chair, secretary,
treasurer, assistant treasurer), as well as at least two county committee members (the number is
based on the town’s vote for the party’s gubernatorial candidate at the last election, not on any
choice by the party). The county committees are to elect their own five officers as well as at
least two delegates (one male and the other female) to form, along with the county chairs, the
state committee.
12
In Vermont, tests of party loyalty or ideological compatibility for admission to a town
caucus are specifically prohibited by law, although the law does, at least, limit each voter to
participation in only one party’s caucus. Obviously, verification that only the voters of a party
participate in its town caucus is impossible in the context of a secret ballot.
13
In the United States, partisan registration is generally regarded as the equivalent of party
membership, and since it is—at least in those states that use “closed primaries”—the criterion for
admission to participate in the selection of party candidates and officials, it satisfies at least part of
the Katz and Mair (1992b: 4) definition of membership. On the other hand, however, partisan
registration entails no obligations to the party, is not subject to party approval, and is generally
administered by the state rather than by the party.

rights to the fulfillment of obligations accepted by states to govern their own
behavior and applied to parties as if parties were part of the apparatus of the
state itself.14
PARTIES AS AGENTS OF CIVIL SOCIETY AND

AS ELECTION CONTESTANTS
One common objective of party laws is to bar from the ballot, or even to allow
the dissolution/banning altogether of groups deemed to be subversive of the
democratic order. More generally, laws might bar advocacy of the use of
violence or other forms of intimidation or appeals to or incitement of racial or
other group hatred. These restrictions might, however, reasonably be imposed
on any organization.
Another major impetus to the adoption of party laws is, however, unique to
political parties in their capacity as sponsors of candidates for election. Once
the state began to provide official ballots for elections, the question of eligi-
bility for inclusion on those ballots had to be resolved, and once party labels
as such were included on the ballot, the question of who was entitled to use a
party’s name—and indeed the question of what organizations were entitled to
have a name that would benefit from this kind of “trademark protection”—
had to be addressed. In particular, the state (the law) had to specify the
requirements/conditions to be recognized as a party, whether all parties
would automatically be entitled to a place on election ballots and to other
privileges (for example, tax advantages) or whether they must satisfy add-
itional conditions, and how the legitimacy of a claim to be the candidate or list
of a particular party would be assessed.
Requirements both for recognition, and for obtaining or retaining a place
on the ballot and other advantages, generally are based on the size or support
of the party—or in some cases on the party’s willingness to put a sum of
14
Examples of such international agreements include the International Covenant on Civil and
Political Rights, the European Convention for the Protection of Human Rights and Fundamental
Freedoms, and the Convention on the Elimination of all Forms of Discrimination against Women
(CEDAW)—which was the basis on which the European Court of Human Rights ruled in the case
of Staatkundig Gereformeerde Partij against the Netherlands that the state is obligated to ensure
that political parties allow women to become candidates, notwithstanding sincerely held religious
beliefs that the “participation of women in both representative and administrative political
organs” is “incompatible with woman’s calling” (par. 9). Even more directly, art. 4 of the
French Constitution obligates parties to “contribute to the implementation of the principle set
out in the second paragraph of article”—to “promote equal access by women and men to elective
offices and posts as well as to positions of professional and social responsibility.”

money at risk in the form of a deposit to be refunded only if that party attracts
sufficient support at the subsequent election. The fundamental questions are
the status to be accorded to small parties (parties with unpopular ideologies or
simply appealing to a very narrow interest), and the height of the obstacles to
be erected against the entry of new parties into the electoral marketplace. The
“public” justification for high barriers to entry is generally the same as for an
electoral system with a high threshold of representation (a strong electoral
system, in Sartori’s (1986) terms): to discourage the fragmentation of the
party system,15 as a result either of the entry of new parties or of the splitting
of old ones, and to protect the political arena from being used by groups that
have no serious interest in participating in political debate.16 Alternatively,
high barriers to entry may really be intended to protect the already recognized
parties from unwanted competition, and to raise the cost of the “exit” option
(Hirschman 1970) for internal dissidents. Similarly, while provisions allowing
adjudication of rival claims to the use of a party’s name may be justified as an
administrative necessity, they may also legislatively fix power relations within
the party.17
PARTIES IN GOVERNMENT
Once candidates have been elected, they have a dual role. On the one hand,
they remain party (wo)men, but on the other hand they are also public officials.
A natural question is the degree to which the parties that occupy public offices
will be able to use the resources of the state for their own ends. Here there are
two major categories to consider. One concerns patronage in the allocation of
15
In the case of Yumak and Sadak v Turkey (2008), for example, the Grand Chamber of the
European Court of Human Rights ruled that the state’s interest in facilitating government
formation in parliament could justify the imposition of a 10 percent threshold for representation.
The state might also be seen to have an interest in promoting clarity at the electoral level—
avoiding situations like the 1997 SMP election in Canada in which more than 16 percent of the
seats were won with less than 40 percent of the vote, the 1993 Sejm election in Poland (PR with a 5
percent threshold) in which fragmentation meant that over one third of votes were cast for lists
than won no seats, or the 2002 presidential election in France (two-round majority) in which
fragmentation of the left-wing vote allowed the candidate of the Front National to advance to the
second round—in which he won less than 18 percent of the vote.
16
Canadian examples of “parties” that might legitimately be denied privilege as described by
Court of Appeal Justice Robert Blair in the case of Longley v Canada 2007 ONCA 852 at para. 54
would include those “interested only in satirizing the political process (the Rhinoceros Party), or
in using the process to promote their commercial interests (the Natural Law Party).”
17
For example, the authority of Canadian party leaders to veto local nominating decisions is
based on sect. 67(4)(c) of the Canada Elections Act, rather than merely being based on party rules.

jobs or public contracts. Once widely accepted as normal (“to the victors
belong the spoils,” as New York Senator William Marcy is reputed to have
said after the 1828 election of US President Andrew Jackson), by the late
nineteenth century the “spoils system” had come to be seen in many places as a
form of corruption. One manifestation of this shift in attitudes was civil service
reform—although low-level jobs in the public service remain a significant
incentive to party activity in some places.
The same kind of questions arise with regard to the other category of
concern—policy. Even if there are policies that are in the long-term public
interest in some objective sense, there are no policies that are perceived by
everyone to be optimally in his or her own immediate interest. Unless electoral
campaigns are totally devoid of policy content, government policy will nor-
mally be skewed toward the preferences of those groups that supported the
parties in power. But while the motivation can be the pursuit of ideologically
informed policy goals that happen to advantage supporters (at least for some,
those advantages may, after all, have been the original motivation for their
support), policy may also be shaded so as to reward identifiable supporters,
with its ideological basis provided as rationalizing cover. Policy also may
simply be the pay-off for benefits given to (or expected by) the parties, or by
the parties’ policy makers in their private capacity.18 While outright bribes,
and the conversion of public resources for the private use of those doing the
converting (roughly, embezzlement), clearly are illegitimate, beyond this the
question largely hangs on one’s understanding of politics. If it is about
the discovery of the public interest, then a politically neutral administration
is appropriate, and the partisan targeting of benefits is at least questionable.
If, however, politics is about building coalitions among conflicting interests,
then patronage, log rolling, and the targeting of club goods and side pay-
ments may be integral to the process.19 All of this returns one to the question
raised in Chapter 1 of whether an election should be seen as a case of voters
hiring parties to govern as agents in their (the voters’) interest, or as a case of
parties and their candidates using policy promises to hire voters to act as
their agents in voting for them so that they (the parties) can enjoy the private
benefits of office.
18
The complexity of this is illustrated by the so-called UCLA model of American parties as
long coalitions of intense policy demanders who then recruit candidates who will be amenable to
the, frequently self-interested, policy preferences of those policy demanders (Bawn et al. 2012).
19
From the perspective of methodological individualism, Buchanan and Tullock (1962) argue
that only Pareto optimality can justify public policy, and that this can legitimately be achieved
through the targeting of benefits to groups who would otherwise be made worse off by policies
that are beneficial at the aggregate level, but not beneficial to all individuals. They ultimately
recognize that the transaction costs of requiring unanimous consent would be prohibitive, and
compromise by suggesting that supermajorities should be required for decisions that have the
potential to impose significant costs on some groups.

The preceding paragraphs have implicitly assumed a winner-takes-all
mentality: the governing party or coalition controls the jobs and controls
policy, and can use these either for their own advantage or in the (presumed)
public interest, while the opposition parties watch, complain, and hope to
become the government after the next election. There is, however, an alter-
native model, that of share-and-share-alike. This was the practice described
by Lijphart (1968) in his model of consociational democracy in the
Netherlands—and later (Lijphart 1999) presented as a contrast to the major-
itarian system of the UK or New Zealand. Well into the post-war era, for
example, it was said that a party card was a requirement for a job cleaning the
trains in Austria, but the card did not have to come from the party currently in
office; according to a poll of party members in the 1980s, gaining access to
public housing was an important or very important reason for membership
for roughly one third of those responding (Müller 1989: 339). The practices
of Proporz in Austria, lottizzazione in Italy, or consociationalism in the
Netherlands all involved allocation of patronage (contracts and control over
public institutions, as well as jobs) to all of the major parties and their
supporting penumbra of organizations, not just those that happened to be
in power at any particular time. The problem of access to state resources
giving an unfair advantage to some parties was apparently solved by giving
access to all (the major) parties, effectively erasing the category of loser.
The share-and-share-alike approach, at least as a tendency, can extend to
systems that are otherwise far from the consociational (or “consensus” in later
elaborations) model. Allocating significant resources to MPs means that all
parties that elect members receive a slice of the cake—generally formally tied
to their public rather than their partisan role, although given that one of the
best routes to re-election is to be perceived by the voters to have done a good
job in office, this distinction is hard to maintain. Parliamentary committee
chairmanships and assignments as rapporteur may be divided among all
parties (e.g., Hungary, Belgium, Finland, and the European Parliament),
rather than being monopolized by the parliamentary/legislative majority
(e.g., the United States or Italy).
Moreover, rather than simply giving resources to MPs as individual mem-
bers, money and staff may be allocated to parliamentary parties per se.20 Once
20
The expansion of resources available to non-government parties in the Canadian House of
Commons provides an example. The leader of the opposition has been paid a special salary (now
equal to that of a cabinet minister) since 1905 (plus a car allowance since 1931 and an official
residence in Ottawa since 1950). Since the 1970s, however, there was an explosion in the number
of non-government party MPs receiving special salaries and allowances to include in 2006: the
leader of the opposition, the leaders of all other parties, the opposition house leader, the house
leaders and deputy house leaders of all other parties, and the whips (both chief and deputies) of all
parties. While of course justifiable in terms of the importance of these offices to the smooth
functioning of Parliament, these salaries also serve to mitigate the costs of electoral defeat (leaders

these party-based advantages are given, however, it becomes necessary to set
some floor on the size of party that is eligible for them, and this, in turn, will
have an important bearing on the capacity of small parties to operate as
independent entities as well as on the costs to schismatics of leaving an
established party group.
As already observed, the earliest constitutional references to parties were
intended to insulate the civil service from partisan intrusion. These were
both preceded and followed by “civil service reforms” with the same inten-
tions. Increasingly numerous and increasingly broad regulations have been
designed similarly to limit the conversion of public resources for partisan
purposes, and in doing so have been modeled on standards that would be
applied to state agencies. This then raises the mirror image of the problem
that civil service positions that should be politically neutral will be captured
for partisan purposes: that those occupying positions that “ought” to be
partisan will “rise above” partisanship and take on the mantel of neutrality
in the public interest. That instead of parties “colonizing” the bureaucracy,
the norms of the civil service will undermine the partisan nature of the
parties.21 While this may have some immediate appeal to those who believe
that there is a singular public interest that can be unproblematically
advanced through “evidence-based policy making” (e.g., Head 2010), it
basically aims to take the politics out of government (“reason is clean and
politics is dirty” (Stone 1988: 305)). The result would be to make politicians
more like public service technocrats rather than representatives of particular
opinions or interests. Evidence of this orientation can be found as early as
the 1970s (for example Crossman 1972).
One final set of regulations that reflects the blurring of the line between
party and state service are those laws or parliamentary rules restricting
the freedom of MPs to switch parties during the course of a parliamentary
term. This phenomenon is relatively uncommon in the established democ-
racies, although by no means unknown.22 In many of the newer democra-
cies, however, party switching has been perceived to be a greater problem,
both undermining the stability of governments and preventing parties
from progressing beyond the stage of personal cliques always “for sale to
and whips may expect to go to the somewhat lower supplementary salaries paid to the opposition
or other party rather than to the simple base MP’s salary).
21
An obvious, although now dated, example of this was the relationship between the Japanese
LDP and the bureaucracy, with senior civil servants. See Thayer (1969).
22
Belinda Stronach’s 2005 defection from the Canadian Conservatives to the Liberals in 2005,
which allowed the Paul Martin minority government to remain in office, would be an example.
Between 1996 and 2001, “almost one-fourth of members of the lower house in Italy . . . switched
parties at least once” (Heller and Mershon 2005: 546), but the Italian party system was at that
time clearly in a state of flux.

the highest bidder” among would-be prime ministers. In a significant num-
ber of these cases, there has been recourse to legislative action to enforce
party stability.23
While these rules raise important questions of democratic theory, their
significance here stems from the prior question of how enforcement of party
discipline becomes a matter for legal regulation in the first place. Similarly,
why are parliamentary parties and not just individual MPs legitimate recipi-
ents of public funds? The most reasonable answer to both these questions is
because parliamentary parties have come to be seen as part of the public
institution of parliament, rather than merely as private associations of MPs
that aid them in coordinating their activities within parliament.
PARTY FINANCE
In the classic days of the mass party, parties were assumed not only to be
separate from the state, but also to be self-supporting. Politics also was
presumed to be labor intensive, with the labor provided primarily by party
members or members of affiliated organizations such as trade unions. Over
the course of the last fifty years, however, this has changed substantially.
Where parties used to depend on their local organizations to keep informed
about public sentiment and to communicate their views to the public, they
now rely heavily on focus groups and public opinion polls for information
about the public, and similarly rely heavily on direct mail (either paper
or electronic) or broadcast, satellite and cable media, and the internet to
communicate to the public. This requires that they deploy not armies of
volunteers, but squads of professionals in survey research, advertising experts,
media consultants, web designers, and so forth. Some of the other implica-
tions of this transformation were discussed in Chapter 3, but here only one
fact needs to be emphasized—this professionalization and mediatization of
politics has been expensive. In what has become a nearly universal response to
the fact that the cost of politics has risen far more quickly than the ability of
parties to raise the necessary money through their traditional sources of mem-
bership subscriptions and contributions from organizations (prominently includ-
ing corporations, trade unions, and interest groups) in civil society, parties have
turned to the state for financial support—not simply to help party (wo)men
23
According to Janda (2009), 14 percent (five [sic]) of older democracies (India, Israel,
Portugal, Trinidad & Tobago), 24 percent of newer democracies, and 33 percent of semi-
democracies have laws against parliamentary party defections.

perform their public duties as holders of public office, but to support the poli-
tical activities (especially campaigning) of the parties as parties.24
This turn to the state has been justified in two ways. On one hand, it has
been supported as a way to limit the influence of private money in politics;
whether or not it was corrupt either in the formal sense of being illegal or in
the more substantive sense of political influence being exchanged for cash, the
large-scale flow of private money into party coffers certainly gave the appear-
ance of corruption, fueled by periodic scandals that left no doubt about the
corrupting influence of money (see, for example, Ames 2014). Even in the
absence of a specific quid pro quo, the need constantly to raise money must, at
a minimum, make parties and politicians disproportionately sensitive to the
interests and concerns of those who have given in the past, and who might be
persuaded to donate in the future.
On the other hand, public support has been justified by the claim, already
articulated above with regard to party laws, that political parties perform
functions that are essential to the maintenance of democracy. Moreover, it is
not just that competition among parties has become the defining characteristic
of modern democracy, and therefore that the state might have an obligation
to assure the survival of parties. In addition, some of the responsibilities of
parties—policy research or civic education, for example—might be regarded
as explicitly governmental functions that should be publicly financed. But as
with support for an MP’s constituency office or a minister’s speech writer, the
distinction between support of a governmental function and support of par-
tisan political activity may be more clear in theory than it is in practice.
Public subventions to parties basically fall into three categories. Generally
the oldest is indirect support in the form of free or subsidized provision of
goods and services, most significantly free access to the broadcast media. The
second, and also indirect, form of subvention falls under the heading of “tax
expenditures”; no money is transferred from government coffers to the par-
ties, but by foregoing (in the form of tax deductions, credits, or exemptions)
revenue that it would otherwise have collected, the government either makes
contributions to parties less costly to the donors, or makes the money raised
by the parties go farther. Finally, and generally of the most recent advent
24
The UK represents a partial exception to the general practice of state subsidy, although
even there parties are eligible for Policy Development Grants totaling £2 million per year, and
opposition parties receive “short money” payments (from April 2014 equal to £16,689 for every
seat in the House of Commons plus £33.33 for every 200 votes gained by the party, equaling
more than £17 million for the Labour Party based on the results of the 2017 election). Although
Canada eliminated per vote subsidy to parties (introduced in 2004 and eliminated in 2015), it
continues to subsidize parties through partial reimbursement of campaign expenses and through
tax expenditure.

(pioneered in Costa Rica in 1954 and Argentina in 1955, and then first
introduced into Europe in Germany in 1959 followed by Austria in 1963
and Sweden in 1966), there can be direct cash payments to the extraparlia-
mentary party organizations. Especially from the perspective of citizens, this
is the most obvious form of public support, and in many cases it is also the
largest. According to data from the Political Parties Data Base project (data
from 2010 to 2013), state subsidies provided more than two thirds of central
party income in nine of the eighteen countries considered, and were the largest
single category of income in five more (van Biezen and Kopecký 2017).
There are many organizations other than political parties that are subsid-
ized by public funds—hospitals, museums, and universities to name only a
few. Although many of these are structured as formally private entities (for
example, their employees are not regarded as part of the civil service), there
comes a point at which their dependence on the public treasury becomes so
great that they effectively become public institutions, with implications for
their mission as perceived both by others and by themselves, and accompany-
ing state intrusion into the management of their affairs including requirements
of transparency and procedural regularity far exceeding what would normally
be required in the truly private sector. Moreover, in some cases the receipt of
public funds has obliged these private-sector institutions to discharge what
would otherwise be explicitly public functions.25 The question is whether, and
at what point, the same principle applies to political parties as they become
more and more dependent on public subventions.
Indeed, the introduction of public subventions for parties has almost always
been accompanied by, and used to justify, extensive regulation of party
financial arrangements. Even before the introduction of public subventions,
however, parties were often subjected to regulations that were far more
intrusive than those imposed on other organizations. Aside from common
prohibitions on the acceptance of contributions from unions, corporations, or
foreign sources—bars that would never be imposed on public museums or
universities, for example—parties might be required to make public disclos-
ures at a level of detail far exceeding those imposed on other formally private,
but effectively public, entities. With the introduction of direct public pay-
ments, these regulations have generally become even more restrictive.
Reliance on the state as a (the) principal source of party finance clearly has
the capacity to alter the balance of internal influence among, and the rela-
tionships of the party with, its more traditional sources of funding. These
changes are discussed elsewhere in this book. Here, however, we underline
two conclusions. First, the nature of public subvention schemes as laws of
general application means that to the extent that they are dependent on
25
For example, the policing of sexual misconduct at American universities.

state resources, all of the parties are operating under the same constraints and
incentives; rather than some parties depending on members, others on
business organizations, and still others on unions or other organizations, all
have the same paymaster. Second, both their financial dependence and the
requirements imposed further the entanglement of parties and the state, and
reinforce the sense that parties have become, at least, semi-state agencies.
PARTY LAWS AND PARTISAN SELF-INTEREST
Most treatments of the institution of party laws take parties as the object, and
some other entity like parliament as the subject: party laws are imposed on the
parties by the state. When regulations are imposed by agencies like independ-
ent constitutional courts, there is some validity to such an orientation. But for
the most part, when laws are enacted, it is by parliament, which given the
nature of parliaments is to say that they are enacted by the parties themselves.
While it may appear obvious why parties would vote themselves large sub-
ventions, it is not obvious why they would also impose strong restrictions on
their own freedom in exchange for these subventions. This is even more the
case with regard to such matters as their own internal organizations or
membership practices, for which there is no direct corresponding pay-off.
Why would parties that have the power to enact laws, enact restrictions on
themselves? In response, it is useful to suggest three potential answers. (For an
analogous discussion with regard to the claim that electoral reforms should
occur “when a coalition of parties exists such that each party in the coalition
expects to gain more seats under an alternative electoral institution, and that
also has sufficient power to effect this alternative through fiat given the rules
for changing electoral laws” (Benoit 2004), see Katz 2005.)
The first answer, particularly applicable to the insurgents in the new dem-
ocracies of post-communist Europe, but also, for example, to the progressive
reforms in the United States in the late nineteenth and early twentieth cen-
turies, is that the parties enacting these laws do not really think of them as
applying to themselves, because they do not (yet) think of themselves as
governing parties, rather than as reformers who are creating a political system
that will be run by others. From this outsider perspective, their partisan self-
interest is assumed to be short-lived, and they are more concerned with
limiting possible abuses by those who will come after them than they are
with empowering themselves.
The second potential answer is that the parties perceive the regulations to be
the lesser of two evils. They recognize that the reforms they are enacting are
not unequivocally in their own interest, but they fear the alternative of doing

nothing even more. Many reforms, for example the party finance reforms
enacted in 1993 by the Liberal government in Canada, can be understood as
regulation for the purposes of damage control. While some reforms are
enacted by parties brought to power in the wake of scandalous behavior by
the parties they have ousted, others are enacted by those caught up in scandal
themselves as evidence that they intend to sin no more. Moreover, even when
reforms are proposed by parties not themselves immediately implicated in
scandal, they may well be supported by those who were implicated in the hope
that such support will be taken as evidence of contrition.
Finally, regulations that do not appear to be in the immediate self-interest of
parties in power—for example, regulations that limit the patronage capacity of
governing parties—may in fact be in their interest over a longer period, some
of which they must expect to spend in opposition. In this case, party laws may
be akin to workplace safety legislation from the point of view of a well-
intentioned employer: investing in a safe workplace is in the employer’s interest
only if (s)he does not risk being undercut by other employers who skimp on
workplace safety. As Olson (1965) suggests, a rational egoist may want to be
coerced to accept restrictions (or to invest in the production of public goods)
that are in his interest only if the restrictions are observed or the costs paid by
everyone, because the coercion makes universal acceptance credible.
Moreover, not all regulations that are general in their application are also
equal in their impact. To continue with the workplace safety analogy, it is easy
to imagine a requirement for the installation of safety equipment so expensive
that it drives small firms out of business. Similarly, many of the provisions of
party legislation in fact work to the relative advantage of larger and estab-
lished parties, and to the detriment of smaller and newer upstarts. To take a
simple example, a newspaper advertisement costs the same amount, and
reaches the same audience, whether it is placed by a large party or a small
party, but this means that if parties are subsidized in proportion to their size,
the larger parties actually are advantaged because they can afford more
advertisements. In this case, the restrictions of party laws may be seen as
providing a form of credible commitment, and political camouflage, that
allows the larger parties collectively to pursue what in fact is in their collective
self-interest of squeezing their smaller challengers. As Bowler et al. (2001)
show with regard to electoral laws, even if cartel parties are unable to entirely
shut out would-be competitors, and indeed even if rules appear to be evolving
towards a more liberal environment for all political parties, which on the
face of it suggests an easier ride for small and/or new parties wishing to
break into the system . . . these developments can have another interpret-
ation . . . suggestive of a cartel-type situation with a twist: nest-feathering
and liberalizing of electoral laws to the benefit of all parties, but is
proportionality more so for the established parties than for the new and/
or small parties.

EXPANSION OF PUBLIC-PARTISAN OFFICES
One of the practices frequently associated with the origins of party govern-
ment was the widespread use of patronage. The party (or parties) in office
would place their supporters not only in policy-making positions, but
throughout the government payroll—as clerks, street sweepers, postmen—as
a reward for party service. And, as a result, a change in government could
mean a wholesale replacement of these party loyalists with the supporters of
the new governing parties.
This aspect of party occupation of the state was to a great extent eliminated
by the establishment of permanent civil services, recruited on the basis of
“merit” (generally defined by test scores or educational achievements supple-
mented by professional experience), and given security of tenure in exchange
for political neutrality. While many of these public jobs have been removed
from the partisan sphere (not least because their relative attractiveness had
declined—see, for example, Sorauf 1959; Kopecký and Scherlis 2008), by the
1970s and 1980s, a large number of jobs for which party ties appeared to be
prerequisites were being created.
One obvious growth area for political appointments is members of minis-
terial cabinets (people appointed by and working for an individual minister).
In Belgium, for example, in 1947 there were 107 cabinet members in the top
pay grade; in 1965, there were 200; in 1986 there were 530 (Brans and
Hondeghem 1999: 131). A similar, albeit smaller and less formal, trend was
evident in Denmark, where the number of ministerial private secretaries and
press secretaries rose from twenty-three and one in 1985 to forty and nine in
1995, respectively (Jensen and Knudsen 1999: 244). In Austria, the number of
members of cabinets ministériels dropped from thirty-two in 1970 to twenty-
eight in 1971, but then rose to sixty-four in 1988 and eighty-eight in 1994
(Liegel and Müller 1999: 101–2).
Another, and in numerical terms often larger, expansion came in the form
of positions in “quangos” (quasi-autonomous non-governmental organiza-
tions) or NDPBs (non-departmental public bodies), or on the boards of public
corporations. Moreover, these new “public-partisan” positions generally are
distinguished from the old patronage positions by the higher levels of profes-
sional qualification required and in effectively being policy making (or at least
policy influencing)—in other words, by being part of the governing establish-
ment and not just rewards for loyalty.
Although not all appointments were in the direct gift of ministers, and
although attachment to the party of the appointing minister was not always
required, the British case represents a reasonably well-documented example.
According to Cabinet Office figures, in 1988 there were more than 2000
NDPBs, representing in aggregate more than 50,000 appointments (Research

Paper 96/72, June 14, 1996); if “local public spending bodies” are included, by
1994 the number of executive quangos had exceeded 5500 (Weir and Beetham
1999: 202). On a smaller scale, by 1993 there were 545 national-level quangos
(selfstandige bestuursorganan) in the Netherlands (Van Thiel 2004: 169).
O’Malley et al. (2012: 216) cite research from the Irish Fine Gael Party that
estimated that in 2008 “there were 445 national state agencies, over 200 of
which had been created in the previous ten years.”
In many quarters, these developments were criticized as reward-for-loyalty
patronage—“jobs for the boys,”26 with an implicit assumption that people
appointed on the basis of partisanship would be incompetent while those
appointed on the basis of professional qualifications would be above politics.
At the same time, however, there was growing recognition that clientelistic
reward for support was not the only function served by the expanded number
of appointments, that “[p]atronage in this sense is . . . a strategy to build
parties’ organizational networks in the public and semi-public sphere. Patron-
age in this sense can best be considered as (one of the forms of) party-state
linkage(s), rather than as a party-society linkage” (Kopecký and Mair 2012:
7–8, citing van Biezen and Kopecký 2007).
Although Kopecký and Mair (2012: 11) emphasize the importance of
patronage as an organizational resource, and suggest that this is indicative
of “the shift in the party centre of gravity from society to the state,” they
largely retain the idea that even if professional connections and qualifications
have become more important relative to pure partisan loyalty in the appoint-
ment process, patronage is particularly manifest when ministers appoint
loyalists of their own party. In doing so, they perhaps underplay the import-
ance of another aspect of the new patronage: even when appointments are
made by partisan office holders (e.g., ministers) and political connections are
required, these positions frequently are open to adherents of the parties that
are not in power at the time rather than being exclusive perquisites of the
parties currently in government—in some cases because the incumbents of
these positions are not replaced every time the government changes and so
may be holdovers from when “their” parties were in office, and in others
because either shared expectation or law do not allow appointments to be
limited to political allies. As the parties become more entangled with the state,
the sharing of appointments among parties both in and out of government at
any particular time, once thought to be a peculiar characteristic of consoci-
ationalism, has become more common.
Early examples of this “share-and-share-alikeism” come from parliaments.
Precisely because parliaments are so overtly partisan, this trend is especially
26
“Tory peers pile aboard the quango gravy train” (Observer, February 6, 1994), or the charge
that under a Labour government 180 appointments were held by thirty-nine members of the TUC
General Council (Skelcher 1998: 85).

unmistakable in that venue. The expansion of supplementary salaries paid to
Canadian party leaders and whips cited above provides one example. Another
is the large staffs of American congressional committees—allocated both to
the majority and to the minority. While, of course, these are justifiable on the
ground that these offices are necessary for the smooth running of Parliament,
that is also to say that they are now understood to be public and not just party
positions, and correspondingly, that the parliamentary parties themselves are
understood to be public institutions. A less formal manifestation of the same
phenomenon is the inclusion of party leaders—particularly leaders of parties
that are out of government—in what would otherwise clearly be government-
to-government interactions.27 Again, this indicates that the parties that are
out of office are nonetheless to be regarded as part of the state writ large.
While the parliamentary cases represent a sharing of publicly funded posi-
tions among the parties, each party still selects the individuals who will fill its
share of the jobs. In other cases, governmental functions are delegated to
boards or commissions on which parties are assured representation, but not
the right to name “their” representatives. The most obvious examples are the
“alphabet agencies” of the United States (e.g., the Federal Trade Commis-
sion, the Federal Communications Commission) whose membership is div-
ided between Democrats and Republicans regardless of the partisanship of
Congress or the president (who appoints all of the members, subject to
confirmation by the Senate which may, or may not, have a majority of its
members from the president’s party).28 The British Audit Commission (which
oversees the finances of local authorities) in 2010 included three local coun-
cilors among its members: one from each of the three major parties. “In
France, the first monetary policy committee of the independent Bank of
France included socialist leaders (Michel Sapin) and Gaullists (Denize
Flouzat)” (Lebaron 2000). In writing about appointments to top positions
in the Danish civil service, Christensen (2006: 1006) reports that “in several of
[the cases in which a person with either a party-political background or a
career in an interest organization has been appointed as agency head] the
27
November 2008 G20 financial summit, ‘Obama names Former Secretary of State Madeleine
Albright, a Democrat, and former Republican Rep. Jim Leach to hold talks with foreign
delegations on his behalf ’ (www.reuters.com/article/idUSTRE4A95CG20081112).
July 2008: ‘A 7-member select US Congressional Delegation from the Republican and
Democratic parties currently visiting Liberia has reaffirmed the commitment of the U.S.
government to the socio-economic recovery of Liberia’ (http://mnyenpan18.blogspot.com/2008/
07/liberian-govt-gets-congressional.html).
28
Strictly speaking, the only requirement is that no party have more than a bare majority on
each of these commissions. In the language of the Communications Act of 1934: “The maximum
number of commissioners who may be members of the same political party shall be a number
equal to the least number of commissioners which constitutes a majority of the full membership of
the Commission.”

T A B L E 5 . 1 UK appointments and reappointments to executive non-departmental
public bodies and National Health Service bodies
Declared political activity
Conservative Labour Other
1996–7 (Cons until May 2, 1997) 5.9 3.3 1.1

1997–8 (Lab) 2.5 14.0 2.2
2000–1 (Lab) 3.9 11.7 2.6
2005–6 (Lab) 2.1 8.3 2.9
2010–11 (Con) 2.0 5.4 3.0
2015–16 (Con + Lib Dem) 6.2+2.2 5.0 1.7
Note: 1996–8, executive non-departmental public bodies and National Health Service bodies; thereafter
advisory NDPBs and “other” included as well.
Source: Reports of Commissioner for Public Appointments
appointing minister selected a person from the opposition or from interest

organizations not belonging beyond his own political base of support.”
Even when partisan inclusiveness is not required, it has become more
common—even in archetypically majoritarian Britain, where the public criticism
of apparently excessive partisanship cited above led to the creation of the Office
of the Commissioner for Public Appointments (OCPA) in 1995.29 Table 5.1
shows the pattern of appointments and reappointments to executive NDPBs
and National Health Service bodies by the “declared political activity”30 of the
person (re)appointed. The table shows a number of things. First, at least if one
assumes that before the reform the vast majority of these appointments would
have gone to people who would have declared political activity (the complaints
would lead to an expectation of close to 100 percent), the reforms were successful
in opening up these positions to people who were not simply party hacks, but at
the same time party activity was no bar to appointment provided a basic hurdle of
competence/qualification was cleared. Second, there was a clear tendency
(2010–11 being the only apparent exception—and that may have reflected
reappointments) for the party in power to appoint more of its own “activists”
than “activists” of the other party. But third, and most significantly here, that was
29
The creation of the OCPA confirms what was said above about regulation and partisan self-
interest. As Flinders and Matthews (2012: 352) observe, “In many respects, the creation of the
OCPA in 1995 was a rational act for the outgoing Conservative government, enabling it to use a
scandal of its own making to constrain the appointment powers of its successor.”
30
Quoting from the reports of the Commissioner for Public Appointments, “Examples of
relevant political activity include standing for political office, acting as a political agent or
canvassing on behalf of a political party. In addition, it also includes making a recordable
donation to a political party totalling more than £5,000 in any calendar year, or more than
£1,000 if made to a subsidiary accounting unit such as a constituency association, local branch, or
women’s or youth organisation.”

only a tendency, with no party monopolizing the appointments made by its
ministers.
Although patronage as a straightforward reward for party loyalty may be
declining, the number of appointments by partisan ministers appears to have
increased. Moreover, as Allern (2012: 288) suggests may be true for Norway,
having political experience and sophistication is increasingly important for
top-level civil servants.31 Particularly as the mainstream parties have grown
closer together in terms of policy (see Chapter 4) so that, as Jensen and
Knudsen (1999: 235) wrote about Denmark, there are “few fundamental
disagreements between parties,” this suggests that the interpretation of polit-
ical appointments needs to be expanded even beyond that of Kopecký and
Mair. Rather than patronage as an organizational resource for a party, it is
becoming an organizational resource for the parties. It is becoming less parties
appointing their own loyalists in competition with the loyalists of other
parties, and more members of a broader political class appointing other
members of the political class because of shared values and orientations that
are not bounded by party.
Taken with the other developments in the relationship between parties and
the state, this also suggests that the idea of parties as networks also needs to be
expanded. On the one hand, modern parties have an existence beyond their
formal organizational structures, including networks of officials in the state
administrative apparatus appointed from outside as well as from within the
formal civil service. One consequence is to blur the distinction between party
and the nominally apolitical, or at least non-partisan, civil service. On the
other hand, particularly where they penetrate into the institutions of the state
and connect all of the mainstream parties to the same administrative appar-
atus, these networks also blur the distinction between one party and another.
31
Egeberg (1998: 10, cited by Allern) concluded that more bureaucrats in the 1970s and 1990s
had occupied central positions in public office before being appointed than had been the case in
the 1930s and 1950s.
The Cartel Party
As we observed in Chapter 1, the processes we have explored in the preceding

chapters are neither new, nor completed. And, if we extrapolate from the trends
that were taken to define party types in the past (the trend to ideological and
social distinctiveness of the mass party, or the “shedding of ideological baggage”
of the catch-all party, for example) they are not likely ever to be fully realized. It
is our contention, however, that the trends we have discussed have proceeded to
such a degree that the traditional ideal types of mass party and catch-all party
are no longer adequate, and a new ideal type is required. We have identified this
new model as the “cartel party” (Katz and Mair 1995, 2009).
In this chapter, we explain why the use of that name is appropriate, and
consider how a reasonable approximation of the ideal type might be mani-
fested empirically. We then assess how the evidence—primarily presented in
the preceding chapters—supports the utility of the cartel party model for
describing and understanding political parties and party politics, particularly
around the turn of the twenty-first century; developments since then are
addressed in Chapter 7. Before doing that, however, we first briefly review
the cartel thesis, particularly as it relates to the evolving relationships among
civil society, parties, and the state—emphasizing that although this treatment
is derived from the experience of Western European parliamentary democra-
cies (for example, claims about “weakening of ties between parties and civil
society” assume that those ties were strong at some earlier time), there are, to
paraphrase a Chinese proverb,1 many paths to a cartel-like situation, and
second, we clarify the status of the cartel party as an ideal type.
THE CARTEL THESIS REVIEWED
The earliest parties were of the caucus or cadre type.2 They developed at a
time in which only a small portion of the population, whether limited by
1
“There are many paths to the top of the mountain, but the view is always the same.”
2
This section draws heavily (and quotes freely) from Katz and Mair 1995.
The Cartel Party 125
Civil society
Party
Party
State
F I G U R E 6 . 1 Parties, civil society, and the state: the caucus party type
Source: Adapted from Katz and Mair (1995)
heredity or property ownership or income, could legitimately participate in

political life. Even when they were not literally the same people, ties of family
and interest meant that there was considerable overlap in personal terms
between those who occupied positions of power in the state apparatus and
the rest of politically relevant civil society. Parties were basically alliances of
individuals within this overlap along with supporters/collaborators outside of
the state. While there would be more than one party, representing different
opinions regarding the public interest (or, less charitably, their private inter-
ests), given the relative homogeneity of civil society (relative to the whole
society writ large), the boundaries between these parties could be quite
porous. And given the relatively small numbers of individuals involved, either
in the parties themselves or in the electorates to which they would have to
appeal, the party organizations themselves could be quite informal. This
pattern is illustrated graphically in Figure 6.1.
As societies became more complex, particularly through industrialization
and its attendant urbanization, three significant changes to this model
occurred. First, they brought about a much clearer separation between the
state and now larger civil society, much of which had no personal connection
to those managing the state, and were inclined to view the state as “them”
rather than as part of “us.” Parties were understood to be providing a
linkage between civil society and the state (Lawson 1980, 2011). Second,
the increased differentiation and segmentation of civil society also increased
differentiation and segmentation of individual parties, making the bound-
aries between them much less permeable.3 Third, the larger scale of electoral
3
This is not to suggest that the cleavages that underlay party divisions only arose at this time
(although that is, of course, true for the cleavage between proletariat and bourgeoisie), but rather
that until the rise of mass suffrage those “sides” that had lost in the state-building process (see
Lipset and Rokkan 1967) were largely excluded from party politics. Moreover, even if the
overlaying of a workers/owners cleavage through the enfranchisement of the working class had
Party
Civil society State
Party
F I G U R E 6 . 2 Parties, civil society, and the state: the mass party type
politics required far more elaborate organizations, not just to coordinate

activity, but especially in the case of parties representing newly enfranchised
citizens, educating their supporters, and integrating them into their political
community (Duverger 1959 [1951]: 63). This gave rise to the mass party and,
if all parties were to conform to the mass party ideal type (which, of course,
was not the case) it would generate the schematic pattern illustrated in
Figure 6.2. The democratic theory of the mass party was also the basis for
the principal-agent model of party government, which explains the similar-
ity between Figure 6.2 and Figure 1.2.
The mass party model of competition through the mobilization of a well-
defined classe gardée obviously had little appeal for the leaders of the established
caucus parties, if only because their natural clienteles were too small to be
competitive on their own; they therefore tried to compete across social segments
by continuing to appeal to an overarching national interest, in contrast to the
mass parties appeal to the interests of its classe gardée—even if sometimes trying
to equate that class interest with the national interest. The normative ideal of
control by the party’s base also had little appeal for those already in charge of
caucus parties, who would be happy to have organized supporters, but were far
less anxious to yield authority to a new set of masters. Significantly, the mass
parties came to see themselves as realistic aspirants to ministerial office, and even
more once they came to power, the advantages (or even the necessity) of freeing
themselves from excessive dependence on their external party organizations, and
the importance of serving interests beyond those of their classes gardées, became
apparent to the leaders of mass parties as well. The result, from both directions,
was the emergence of the catch-all party, which in ideal type form would serve as
a broker among interests/social segments rather than as the privileged agent of
any one of them. But this necessarily involved weakening or abandoning its
in some places the effect of increasing the permeability of cleavages on the bourgeois side of the
divide, the left/right cleavage itself was much less permeable.
Party
Civil society State
Party
F I G U R E 6 . 3 Parties, civil society, and the state: the catch-all party type
Th parties
The rties State
Civil society
F I G U R E 6 . 4 Parties, civil society, and the state: the cartel party type
roots in civil society so that it would have the independence to act as an honest
broker, leading to the pattern illustrated in Figure 6.3.
If these three “snapshots” were taken instead as three frames from a moving
picture, one could imagine three trends emerging. The first, already evident in
Figure 6.2 (i.e., with the mass party), is the separation of state and civil
society. The second, evident in the comparison of Figures 6.2 and 6.3, is the
withdrawal of party from civil society, and its greater entry into the state. The
third, also evident in the comparison of Figures 6.2 and 6.3, is the movement
of the parties toward each other. The cartel thesis projects these trends as
continuing, ultimately (again, in ideal typical terms) with the connection
between party and civil society being largely severed, and with the governing
of mainstream parties becoming so similar to one another in structural
characteristics, policy proposals, personnel types, and self-referential interests
that it becomes reasonable to think about “the parties” as a group, rather than
as individual parties to be considered independently.
But—and it is a crucial “but,” to which we return in Chapter 7—the cartel
model, represented in Figure 6.4, is not a stable end state, nor is it, notwith-
standing one possible interpretation of the presentation above, the result of a
simple linear process. On the one hand, the evolution from caucus to mass to
catch-all to cartel parties has been brought about by the adaptation of party
politicians to changing circumstances, with each adaptation generating a

reaction, leading to a new adaptation; there is no reason to suppose that
circumstances will not continue to change, and politicians will not continue to
adapt. On the other hand, while we know that those in power tend to be
reluctant to change, the evolution of party types was also driven by their own
limitations or internal contradictions. The cartel model is no exception, with
many of its frailties, some highlighted by us in our first cartel party paper
(Katz and Mair 1995) and others identified by critics either of the model or of
the behavior to which it has led, usefully summarized by Yael Yishai (2001).
And if, as we speculated in 1995, the cartel party model is one that prepares
the ground for its own challengers (or that sows the seeds of its own destruc-
tion), the conclusion must be, to quote Heraclitus, that “everything changes
and nothing remains still” (quoted in Plato’s Cratylus).
IDEAL TYPES IN SOCIAL THEORY
The cartel party model, like the caucus party, the mass party, and the catch-all
party models, is an ideal type. Ideal types have two distinct uses. One is to be a
theoretical primitive, used to theorize about relationships and processes in the
absence of the messy complications of the real world. In physics, the “perfect
vacuum” is an ideal type, perhaps closely approximated in the laboratory, but
never realized, even in “the vacuum of outer space.” In the social sciences, the
rational economic man is similarly an ideal type, albeit one that is much less
well approximated—ultimately giving rise to the field of behavioral econom-
ics. The “stylized” models of rational choice theory within political science are
constructed out of such ideal types. While they can be heuristically useful even
if there are no real-world cases that approximate them, these idealizations are
useful empirically only to the extent that they are simplifications, but not gross
distortions, of reality. Notwithstanding its continuing normative appeal, it
seems apparent that the use of the mass party ideal type as a primitive in
theories about contemporary democratic practice is inappropriate precisely
because it is such a gross distortion.
This is not, however, the only way in which ideal types can be useful. All
of the models of party encompass many different attributes or dimensions.
These dimensions need not be orthogonal (that is movement along one
dimension may be empirically, or even causally, connected to movement
along others), but so long as they are not perfectly correlated, they jointly
define a multi-dimensional space. It is, however, a space that is devoid of
landmarks. Another function of ideal types is to provide landmarks through
the use of which the dimensions defining the space, and the relationships
among them, can be more easily understood. From this perspective, the fact

that there are no real cases that closely approximate an ideal type is of less
relevance than the fact that movement toward, or away from it can
be understood.4 For Kirchheimer, the phenomenon of interest was the
transformation of mass parties into catch-all parties (more properly, the
transformation of parties that more closely approximated the mass party
type into parties that more closely approximated the catch-all type): not, for
example, the creation of new and totally non-ideological parties, but the
“[d]rastic reduction of the party’s ideological baggage” (1966: 190) by
already existing mass parties. In our case, the phenomenon of interest is
the organizational and strategic transformation of mainstream parties that
approximate the catch-all type into parties that approximate the cartel type.
Quite aside from the fact that ideal types are never fully realized, meaning
that none of these transformations is hypothesized to be complete, neither
are these transformations hypothesized to be instantaneous. Thus the ques-
tion is not so much whether there are a large number of parties that
are reasonably good approximations of one of the ideal types (however
“reasonably good approximation” is defined), but rather whether parties
are moving closer to one of the ideal types in the ways, and under the
conditions, that are usefully illuminated by the model.
CARTELS IN ECONOMICS AND POLITICS
Political commentators have long found analogies to the liberal economic

market in goods and services appealing in thinking about liberal, particularly
liberal democratic, politics. Although (to our knowledge) neither of them
actually used the phrase, there is general consensus that the metaphor of a
“marketplace of ideas” can be traced at least to John Milton’s Areopagitica
(1644) and John Stuart Mill’s On Liberty (1859). Downs’ An Economic
Theory of Democracy (1957) represents a direct application of economic
spatial models of market competition (e.g., Hotelling 1929; Smithies 1941)
to electoral politics. While no doubt influenced by thoughts of the Athenian
Agora, which was a physical place both for the buying and selling of goods
and where citizens might meet to discuss political ideas and proposals, the
overt use of the marketplace analogy came to particular prominence with the
4
For example, Dahl’s (1971) ideal type of polyarchy, defined as the “upper right” (high values
on both dimensions) of a two-dimensional space defined by Inclusiveness and Liberalization, has
proved useful notwithstanding that no real case is a perfect polyarchy, and that “perhaps the
predominant number of national regimes in the world today would fall into the mid-area” (8).

acceptance of the capitalist free market ideal type as the optimal form of
economic organization.
This is not the place to rehearse economic theory, but because the market
analogy has been, and continues to be, so influential in thinking about party
politics, and because of the obvious derivation of our use of the word “cartel”
from its use in economics, it is useful briefly to highlight a few of the
assumptions that underlie the idea of the free market—and that are necessary
conditions for that market to achieve the optimality attributed to it.
The free market is based on competition by self-interested actors. As Adam
Smith (1937 [1776]) said, “It is not from the benevolence of the butcher, the
brewer, or the baker that we expect our dinner, but from their regard to their
own interest.” This is a basic assumption of classical economics, and indeed of
any theory that assumes human rationality. More importantly, however, a
free market requires that none of the actors have sufficient economic power,
either as suppliers or as consumers, that they can unilaterally affect the terms
of trade. Re-enforcing this, a free market also requires unrestricted entry into
the market, so that if existing suppliers fail to satisfy demand at a price
consumers are willing to pay, new producers can enter to do so themselves.
Although less often acknowledged, the free market also requires a government
that is willing and able to enforce contracts, suppress violence, and protect
against fraudulent practices.5
Whether or not some markets ever approximate this ideal type sufficiently
for characterizing them as “free markets” to be regarded as an empirically
useful simplification, rather than as a misleading distortion, is a debatable
point. What is clear, however, is that real markets can fail adequately to do so
in many ways. One is the excessive granting of monopolies by the state, as was
common in mercantilist Europe (and to a lesser extent is inherent in the
granting of patents and copyright protections). Another is state favoritism in
law enforcement. Of particular relevance to the cartel party argument, even
without state intrusion, the assumption that no actor can unilaterally affect the
terms of trade can be violated in either or both of two closely related ways. On
the one hand, either the supply or demand for a product can become concen-
trated in so few hands that each of those actors can unilaterally affect the
terms of trade; for example, it is estimated that in 2014 roughly 70 percent of
the market in corn seeds in the United States was controlled by just two
corporations, Monsanto and DuPont Pioneer (Agrimarketing 2015). In this
oligopolistic case, the actors are still in competition, but each with sufficient
market power that its competitors must respond to its actions in particular and
not just to anonymous “market forces” in general. Moreover, oligopolistic
5
Rather than assuming state-supplied protection against fraud, the free market ideology might
instead assume perfect information, in which case fraudulent practices (deliberate deception to
secure unfair gain) would be impossible.

competition typically is about market share, and manifested by increased costs
of promotion rather than decreases in sales price—which may even be
increased to support the cost of promotion. On the other hand, even a some-
what larger number of suppliers (or consumers) can agree to form a cartel, and
to allocate market shares by negotiation and joint decision rather than com-
petition. Perhaps the best-known cartel is the Organization of Petroleum
Exporting Countries (OPEC), whose thirteen members controlled over 80
percent of the world’s proven reserves of crude oil in 2015, and about 40
percent of the world’s oil market. OPEC is an intergovernmental organization
with a board of governors and formal decision-making processes, and cartels
generally are assumed to involve formal agreements among suppliers to main-
tain prices and/or restrict competition. The fuzzy boundary between oligopoly
and cartel stems from the possibility of tacit agreement through signaling
(D’Aspremont et al. 1991). Rather than establishing a formal and semi-
permanent organization, or attempting to negotiate a one-off deal with other
oligopolists (both of which would be illegal in most capitalist countries), one
oligopolist may signal to others that he/it wants to raise prices simply by doing
so. If others follow suit, a higher price is established more or less as it would
have been with overt negotiation; if not, the originating actor almost immedi-
ately returns to the old price. This process may be repeated several times within
a short period until a new equilibrium is established.6
There are some respects in which any market analogy must fail when
applied to politics. Most simply, the economic market is about the production
and distribution of private goods (cars, apples, football matches). On the one
hand, each consumer is free to decline to purchase at the price offered. On the
other hand, only those consumers who choose to purchase have to pay, and
only those consumers who pay get to enjoy the goods. The political system,
however, is primarily about the production of public goods (laws, defense,
public heath). While citizens in democracies may be offered a choice of
packages of these goods (generically encompassed in party programs), the
decision is collective rather than individual; electoral winners and losers alike
share in both the costs and benefits of the goods, and no one has the choice of
opting out (except, perhaps, through emigration).7
Against this background, the political analog to the free market in goods and
services has generally been understood as free and fair electoral competition
6
Particularly with regard to the ostensibly competitive domestic market, changes in American
airfares provide a good example.
7
This is obviously an oversimplification, ignoring, for example, externalities (for example, the
costs of pollution generated by manufacturing that are not included in the price charged for the
good produced or paid to those who suffer because of the pollution) in the case of private goods,
and the fact that governments produce private goods as well as public goods (or may “privatize”
what might otherwise be public goods, for example by charging user fees to access public parks).

among independently competing political parties.8 Ultimately, it is through the
medium of this analogy that the principal-agent models of party government
raised in Chapter 1 are constructed and defended. In place of the “discipline of
the market” which is presumed to force producers to be responsive to their
customers, the “discipline of the ballot box” is presumed to force parties to be
responsive to their electors.
If concentration is a problem in the economic market, however, it is
the normal state for the electoral market. The US Justice Department, for
example, considers a Herfindahl-Hirschman Index (HHI) of 1500 (out of
10,000) to indicate a “moderately concentrated” market, and an HHI over
2500 to indicate a “highly concentrated” market (United States Department
of Justice 2010: 19). Translating that into Laakso and Taagepera’s (1979)
effective number of parties (which is simply the inverse of the HHI scaled to
have a maximum of 1 rather than 10,000) would identify any system with
an effective number of parties below 4 to be highly concentrated (and
below an effective number of 6.67—roughly half the number of members of
OPEC—to be moderately concentrated). As Table 6.1 shows (based on the
first election after 2000: for more recent data see Chapter 7), fully ten of the
thirty countries listed had highly concentrated party systems at the electoral
level—a number that rose to seventeen out of thirty with regard to legislative
parties; even at the electoral level, only three (Belgium, Latvia, and Slovakia)
exceeded the threshold of moderate concentration—and given that Belgium
might better be described as having two separate party systems, it is not clear
that it should be included in that list. In other words, even if elections are free
and fair, the political market in most modern democracies is, and effectively
always has been, at best oligopolistic. In this sense, the appropriateness of the
free market analogy was always open to question.
It is our contention, however, that by the late 1990s, the political market
had gone beyond mere oligopoly to be in many countries increasingly dom-
inated by a de facto cartel comprised of all of the mainstream parties—and
indeed that it was complicity in this cartel that had come to define the
mainstream. The most prominent example was, of course, Switzerland,
where the same four parties had formed the Federal Council, in the same
proportions, and with the presidency automatically rotating among them,
continuously since 1959—and continuing to the present, with only a minor
redistribution of council seats from 2003 and a brief hiccough occasioned by a
schism in one of the parties in 2007–8. The “grand coalition” of ÖVP and SPÖ
that governed Austria continuously from 1945 to 1966 was a shorter-lived
example. In both these cases, joint decision making by the entire cartel was
8
For an assessment of the meaning of “free and fair” in this context, see Goodwin-Gill (1994,
2006) and Boda (2004).

T A B L E 6 . 1 Effective numbers of parties
Country Effective number of Effective number of

electoral parties parliamentary parties
Australia (2001) 3.44 2.49

Austria (2002) 3.02 2.88
Belgium (2003) 8.84 7.03
Canada (2004) 3.78 3.03
Czech Republic (2002) 4.82 3.67
Denmark (2001) 4.69 4.48
Estonia (2003) 5.42 4.67
Finland (2003) 5.65 4.93
France (2002) 5.22 2.26
Germany (2002) 4.09 3.38
Greece (2004) 2.66 2.19
Hungary (2002) 2.94 2.21
Iceland (2003) 3.94 3.71
Ireland (2002) 4.13 3.38
Italy (2001) 6.32 5.30
Latvia (2002) 6.78 5.02
Lithuania (2004) 5.78 5.46
Luxembourg (2004) 4.26 3.81
Netherlands (2003) 4.99 4.74
New Zealand (2002) 4.17 3.76
Norway (2001) 6.18 5.35
Poland (2001) 4.50 3.60
Portugal (2002) 3.03 2.50
Slovakia (2002) 8.87 6.12
Slovenia (2004) 6.02 4.90
Spain (2004) 3.00 2.53
Sweden (2002) 4.51 4.23
Switzerland (2003) 5.44 5.01
UK (2001) 3.33 2.17
USA (2002) 2.15 2.00
Source: Michael Gallagher, 2017. Election indices dataset at http://www.tcd.ie/
Political_Science/staff/michael_gallagher/ElSystems/index.php, accessed 24
March 2017
explicitly part of the government formation and governing processes. The

system of consociational democracy that Lijphart (1968) described for the
Netherlands was also an early example of a party cartel, although in that case
not all of the cartel members were in the cabinet at any particular time, and a
shared culture of mutual acceptance and restraint rather than overt coalition,
as well as a sharing of benefits that did not respect the distinction between
those temporarily in government and those temporarily in opposition, char-
acterized this proto-cartel. In the 1960s, arrangements such as these were
considered to be exceptional. By the late 1990s, whether as a result of overt
negotiation, covert collusion, or simple signaling among actors that had come

to recognize and act upon shared interests and constraints, cartel-like behav-
ior if not the formation of formal cartels had to a great extent become the
norm in most of the established democracies.
A CARTEL OF PARTIES
As this discussion suggests, the idea of a party cartel is derived from the idea
of a cartel of producers in the economic market. Within that context, the
nature of those producers, whether individual proprietors, partnerships,
closed corporations, or publicly traded corporations (or, as in the case of
OPEC, sovereign states) makes no difference; what identifies them as a cartel
is only that they act in concert to dominate the market. Maintaining the
analogy to party politics, the existence of a cartel of political parties should
likewise not depend on the way those parties are individually structured.
Although the German term Kartell may simply denote a passing coalition, it
is common to talk about “the Kartell” and “the Kartell parties” (conservatives,
free conservatives, and national liberals—all elite parties) in Germany in 1887
(e.g., Pflanze 1990, 3:232, 350, 392). As one would expect of parties in a cartel,
they agreed not to compete with one another: the other two parties not
contesting districts already held by the third Kartell member; coalescing
behind the strongest candidate of one of the Kartell parties in a runoff
election; and not attacking one another in the press or public meetings.
Without a public statement by the participants that they have formed a
cartel, one can only look for cartel-like behavior by parties as evidence that a
cartel exists. Moreover, it is possible that the behavior observed will only
approximate that which would be expected in a fully fledged cartel. For
example, while the parties in the nineteenth-century German Kartell, the
parties in the 1945–66 Austrian cartel, or the Swiss cartel since 1959 were all
in cabinet office together, in the Dutch consociational system, shifting com-
binations of the cartel parties formed particular governments. Nonetheless,
the first indicator of the existence of a party cartel would be a fairly clear
distinction between two groups of parties, with one group encompassing
parties for which it would not be unreasonable to expect that any one might
go into a governing coalition with any (combination) of the others, and the
other group comprised of parties that would be unlikely to enter government
with any of those in the first group, or indeed to enter into government office
at all. A somewhat weaker indicator would be the apparently permanent
exclusion of some parties from government on the basis of a broad agreement
that they are “beyond the pale,” even if there are some governing formulas
among the other parties that remain essentially unthinkable.

T A B L E 6 . 2 Patterns of coalition formation, 1990–2015
Country Parties in Number of Pairs never occurring

government at possible
least once pairs
Austria SPÖ, ÖVP, FPÖ, 6 (SPÖ, FPÖ), (SPÖ, BZÖ)

BZÖ
Belgium SP, PS, A, E, 36 (CVP, A), (CVP, E), (CVP, FDF),
FDF,9 CVP, (PSC, A), (PSC, E), (PSC, FDF)
PSC, VLD, PRL
Denmark SVP, SD, RL, ZD, 21 (SVP, ZD), (SVP, CVP), (SVP, Kon),
CVP, Kon, V (SVP, V), (Kon, SD), (Kon, RL), (Kon,
ZD), (Kon, CVP), (V, SD), (V, RL),
(V, ZD), (V, CVP)
Finland SKDL, SDP, GR, 21 (SKDL, KESK)
KESK, CHR,
SW, KON
Germany GR, SPD, FDP, 6 (GR, FDP), (GR, CDU), (FDP, SDP)
CDU
Greece PASOK, ND, 6 (LAOS, D’Mar)
LAOS, D’Mar
Ireland DL, LAB, FF, FG, 15 (DL, FF), (DL, PD), (DL, GR), (LAB,
PD, GR PD), (LAB, GR), (FG, PD), (FG, GR)
Luxembourg LSAP, GR, CSV, 6 (GR, CSV)
DP
Netherlands PvdA, D’66, CDA, 15 (PvdA, LPF), (D’66, LPF), (D’66, CU),
VVD, LPF, CU (CU, VVD), (CU, LPF)
Norway SV, A, V, KRF, 21 (SV, V), (SV, KRF)(SV, H)(SV, FRP),
SP, H, FRP (A, V), (A, KRF), (A, H), (A, FRP),
(V, FRP), (KRF, FRP), (SP, H), (SP,
FRP)
Portugal PS, PSD, CDS 3 (PS, PSD), (PS, CDS)
Sweden SD, G, ZE, CD, 15 (SD, ZE)(SD, CD)(SD, LI)(SD, M)
LI, M (G, ZE)(G, CD)(G, LI)(G, M)
UK Lab, LD, C 3 (Lab, LD)(Lab, C)
Evidence concerning this indicator of cartelization was discussed in

Chapter 4, for example the grand coalitions in Germany and Austria, the
magic formula in Switzerland, as well as the fracturing of the pattern of
Fianna Fáil versus the rest when Fianna Fáil and Labour formed a governing
coalition in Ireland in 1992. Table 6.2 looks more systematically at govern-
ments formed between 1990 and 2015 in a range of countries that experienced
at least one coalition government. While some countries maintained a rigid
two-bloc pattern (e.g., Denmark: Conservatives and Liberals versus coalitions
containing virtually every possible pairing of the Socialist People’s Party,
9
Disappeared as a separate list after 1993.

Social Democrats, Social Liberals, Center Democrats, and Christian People’s
Party; Sweden: Social Democrats and Greens versus Center Party, Christian
Democrats, Liberals, and Conservatives), in many of the others what may at
one time have been two or more blocs appear to have merged into one.
In Finland, at the other extreme, all but one of the twenty-one possible
combinations of governing parties appeared at least once; in Greece, trad-
itional antagonists PASOK and New Democracy formed a coalition; in
Belgium and Luxembourg, every possible pairing except Greens and
Christian Democrats occurred at least once; in the Netherlands, except for
the ChristenUnie (which never exceeded 4 percent of the vote) and the
populist Lijste Pim Fortuyn (only in existence for five years), every party
that was ever in government during this period was in government at least
once with every other governing party.
In the ideal type of an economic cartel, one would expect there to be no
overt competition for market share among the cartel members, and indeed the
existence of such competition would signal a shift from cartelized to oligop-
olistic behavior. Notwithstanding the non-competition of the relatively
short-lived—and pre-democratic—nineteenth-century German cartel, the
complete elimination of intracartel electoral competition cannot be expected
in modern democracies. On the one hand, given that the contesting of elec-
tions is one of the defining characteristics of political parties in democracies
and that having elections that are contested is one of the defining character-
istics of those democracies, failure to have at least the appearance of electoral
contestation, rather than indicating a party cartel with full control over the
political market, would simply indicate the destruction of that market. On the
other hand, a durable cartel that was more limited in size but totally precluded
intracartel competition would effectively have become a single party.10 This
does not mean, however, that competition within the cartel cannot be man-
aged and constrained. In particular, if the “product” offered by parties is
policy, cartelization would be indicated by limiting competition over policy
offered by cartel members, reflected both by narrowing the range of options
proposed with regard to any particular policy question and by removing
potentially disruptive, cartel-threatening questions from active competition
altogether. But it is important to note that this constriction of the active policy
space does not represent collusion to impose a particular desired policy, but
rather a tacit “conspiracy” to accept a single position so as not to have to
compete with regard to that policy area at all.
Of particular relevance here is the evidence of the apparent contraction of
the left-right political space as perceived by experts and as derived from party
10
The Australian alliance of Liberal and National Parties is a close approximation of this, with
the parties actually merged into a single organization in Northern Territory and Queensland.

manifestos that is covered by mainstream parties (Tables 4.3 and 4.4). Also of
relevance is the pro-Europe consensus of mainstream parties within the EU,
effectively taking the European issue off of the table. For example, the vast
majority of referenda held within the EU (on British exit from the EU
(2016),11 the Treaty of Lisbon (Ireland, 2008–9), the draft Constitution for
Europe (Spain, France, the Netherlands, 2005), the Euro (Denmark, 2000),
the Treaty of Nice (Ireland, 2001), the Treaty of Amsterdam (Ireland and
Denmark, 1998), the Maastricht Treaty (France, Ireland, Denmark, 1992))
saw all of the mainstream parties in support. That this party consensus did not
simply reflect public opinion is demonstrated by the fact that many of the
referenda were decided by quite narrow margins—and in a number of cases
(perhaps most notably the 2016 Brexit referendum) by the defeat of the
position that the mainstream parties had supported.
The reasons for forming an economic cartel are some combination of
minimization of risk and, in the jargon of modern economics, the extrac-
tion of rents (that is, the reaping of profits in excess of what a free market
would allow). There is no single measure of value in the political market
that is the equivalent of money in the economic market, but the basic ideas
of rent seeking and risk minimization can still be applied. Policy distortion,
that is to say the adoption of policies other than the first preference of the
median voter (which is frequently presumed to be the output of a free
market in policy12—see, for example, Tromborg 2014) might be taken to
indicate rent seeking, but more generally, policy distortion is not the only
form of rent that members of a party cartel might seek, and moreover it is a
form of rent the value of which would depend on the parties as rent seekers
having strong policy preferences that differ from those of the voters. To the
extent (which we have argued is growing) that parties and their leaders are
interested in their own well-being, however, other forms of rent, such as
material (cash) payments to the parties and their core personnel or the
11
The British also held a referendum in 1975 on membership of the European Communities
under renegotiated terms. While the Conservatives supported remaining, Labour formally was
neutral. Prime Minister Harold Wilson and Chancellor of the Exchequer James Callaghan both
supported the “yes” side, although a special one-day conference had voted overwhelmingly to
leave and seven of twenty-three cabinet ministers supported the “no” side.
12
In fact, this presumption, which is generalized from the “Downsian” model of two-party
competition in a unidimensional policy space, is inappropriate in two ways. On the one hand, two-
party competition would mean an oligopolistic rather than a free market, and moreover, with
more than two parties the first preference of the median voter is not necessarily a Nash
equilibrium. For example, with four parties, the equilibrium position is with two parties at the
first quartile and two at the third quartile. On the other hand, as Plott (1967) has shown, in a
multi-dimensional policy space there is unlikely to be any equilibrium position at all, and hence
the expected outcome of free market competition would be indeterminate.

imposition of regulations that tip the competitive balance in favor of cartel
members, and work to the relative disadvantage of parties that are outside
of the cartel, become more important. As discussed in Chapter 5, common
examples of ostensibly neutral regulations that actually advantage larger
parties include ballot access rules or media allocation formulas that advan-
tage incumbent parties, onerous requirements for party registration (for
example, a requirement that a new party have a membership presence
throughout the country), a requirement of expensive cash deposits,
and indeed any allocation of public support that is based on previous
electoral success.
All that said, perhaps the best economic analogy to a cartel of parties is not
to manufacturers of material goods but to the cartel of the teams comprising a
professional sports league. The “products” are entertainment and the emo-
tional rewards of attachment to a particular team, neither of which can be
produced in the absence of competition. The competition on the field is real,
and the winners are rewarded for their victories. But at the same time, the
competition must be managed. If a single team becomes too dominant,
interest falls off—and with it gate receipts and advertising revenue. Thus
one sees devices like team salary caps (common for American professional
sports leagues, but also found, for example, in the Australian National Rugby
League and, in modified form, in the British Premier (football) League) and
sharing of television revenues or simple cash transfers from richer to poorer
franchises, intended, on the one hand, to assure the entertainment value of
competition on the field, and on the other hand to mitigate the economic risk
to the owners of teams that are less successful on the field. Although there may
be rule changes from time to time, these are not adopted to favor one team or
one style of play over another, but to make the game more exciting and
therefore more marketable. From the perspective of the team owners (the
analog of party leaders), in other words, the real question is not who wins on
the playing field but how the economic rewards of the business are divided—
and it is the competition over those rewards that the league (cartel) is formed
to manage. The most direct political equivalent, and thus an additional
indicator of cartelization, would be the kind of sharing of appointments
with nominal competitors, and other measures designed to mitigate the
costs of losing elections, discussed in Chapter 5.13
13
An extreme example of the degree to which this can go was the 1996 appointment of former
prime minister Kim Campbell to be Canada’s consul general in San Francisco. In 1993, Campbell
had led the Progressive Conservatives to a landslide defeat (going from an absolute majority of
154 seats out of 295 to only two seats), but three years later she was appointed consul general by
her Liberal Party successor, Jean Chrétien.

THE CARTEL PARTY
The nineteenth-century German party cartel was made up of elite parties, with
no pretense of internal party democracy or mass membership control. The
party cartel associated with consociational democracy, arguably established
after the resolution of the Dutch school issue in 1917 and reaching its peak
during the German occupation in the 1940s with the Politiek Convent (a
council composed of the two top leaders of each of the then six major parties),
was in many ways a cartel of mass parties, with their associated panoplies of
ancillary organizations, reflecting a segmented social order. But in one
important respect these parties departed, and throughout the life of the
consociational system continued to depart, quite significantly from the mass
party ideal type. At least with respect to coalition policy and thus also with
respect to political policy more generally, they were not internally democratic.
Instead, one of the essential features of the consociational system was elite
autonomy from their members in dealing with the elites of other segments/
parties, precisely so that they could make the accommodations necessary to
maintain the system (Lijphart 1968: 141–6, and esp. 206: “the elites have
usually enjoyed great freedom to act independently without constant
demands from their followers—a freedom vital to the success of the system
of overarching cooperation among the blocs”).
As the experience of the Politiek Convent might suggest, external threats can
have a significant impact on the willingness of erstwhile enemies to reach accom-
modations, and it is perhaps not coincidental that the three classic examples of
consociational systems—the Netherlands, Belgium, and Switzerland—were all
small countries with powerful neighbors, or that the threat of Soviet intervention
may have encouraged cooperation between ÖVP and the SPÖ in the decades
following World War II. Although the post-war settlement reduced the power of
the threat posed by Germany and France to their smaller neighbors to compel
all-party accommodation, the advent of the Cold War and the perceived threat
posed by communist parties oriented to Moscow encouraged accommodation
among social democratic, Christian democratic, and bourgeois parties.
But, as the process of cognitive mobilization (“the increasingly wide distri-
bution of the political skills necessary to cope with an extensive political
community,” Inglehart 1970: 47; see also Dalton 1984) proceeded, and the
weakening of cleavage structures reduced both unity within a social segment
and the willingness of its members to defer to its leadership cadre, the kind of
elite autonomy required to maintain these accommodations became more
problematic within the framework of mass parties everywhere. Combined
with the trends discussed in earlier chapters that have generated a desire (or
need) for the party in public office to secure its dominance within the party
central office, and its independence from the party on the ground, this all leads

to the expectation of a new model of internal party organization compatible
with the demands for maintaining cartel-like relations with other parties: the
cartel party.
Adapting from our original article (Katz and Mair 1995), Table 6.3 sum-
marizes the characteristics that we identified as defining the cartel party type,
and contrasts them with earlier ideal type party models.
One of the core weaknesses of rational choice models of political action is
that rationality can only be defined in terms of objectives, and yet objectives
can only be inferred from action (in which case the exercise runs the risk of
becoming circular), or, alternatively, they must be assumed. Our claims about
the “principal goal of politics” are no different. But if we cannot make
indisputable claims about motivations, we can show patterns of practice.
Much of this evidence was presented in Chapter 3, and suggests that growing
numbers of individuals are able to live from politics rather than for politics.
Even more directly, although limited to a single case, Table 6.4 shows that the
proportion of members of the British House of Commons whose “main
former occupation” was “politician/political organizer” has more than quin-
tupled since the 1960s, and that while this growth has been particularly great
for the Labour Party (grown by a factor of 10), even for the Conservatives
(whose MPs are far more likely to have been lawyers or else company
executives or directors—50.1 percent in 2015 as opposed to 17.2 percent for
Labour), the proportion of professional politicians has more than tripled.
At least in this case, politics has become more professionalized, not in the
sense in which the term is often used in this context to indicate either higher
levels of education or higher rates of recruitment from “the professions,” but
in the sense of being a profession itself.
That party work and campaigning have become more capital intensive is
evident in the increased importance of new—and generally expensive—
techniques and technologies (which even if cheaper per message unit require
large capital outlays and trained operators). Conversely, “the new technology
has been accompanied by the concomitant decline in the importance of old
low-tech methods . . . Canvassing . . . had declined spectacularly even in Great
Britain, its traditional home” (Butler and Ranney 1992: 281–2).
The obvious consequence of capital-intensive party work is for parties to
need more money. Given that party spending is generally limited by party
income—and one of our claims is that party needs have been rising faster than
the willingness or ability of traditional sources to pay—data on the “explosion of
costs” (to use the phrase from Nassmacher 2001) probably understate the
growth of needs. Nonetheless, they have in many cases been quite spectacular.
Van Biezen and Nassmacher (2001: 135) calculate per voter campaign expenses
in Italy to have more than tripled between 1976 and 1987; Blechinger and
Nassmacher (2001: 159) calculate a similar rise in declared Israeli campaign
spending between 1977 and 1988. The response has been a turn to state, for
T A B L E 6 . 3 Characteristics of party ideal types
Elite party Mass party Catch-all party Cartel party
Principal goal of Maintenance of upper Social reformation (or opposition to Social amelioration Politics as a profession
politics and upper-middle it in the case of non-class-based
class dominance and mass parties)
the distribution of
privileges
Nature of party Personal social Labor intensive Both labor and capital intensive Capital intensive
work and networks
campaigning

Principal source Personal connections Members’ fees and contributions Contributions from a wide variety of State subventions
of resources sources
Relations The “elite” are the Bottom up (pace Michels); elite Top down; members are organized Stratarchy; mutual autonomy
between ordinary members accountable to members cheerleaders for elite
ordinary
members and
party elite
Character of Small and elitist Large and homogenous; actively Membership open to all Neither rights nor obligations important
membership recruited and encapsulated; (heterogeneous) and encouraged; (distinction between members and non-
membership a logical consequence rights emphasized but not members blurred); emphasis on
of identity; emphasis on rights and obligations; membership marginal members as individuals rather than
obligations to individual’s identity organized body; members valued for
contribution to legitimizing myth
Basis of claim to Ascribed status Group representation Provision of benefits Efficient management
support
Party channels of Interpersonal networks Party provides its own channels of Party competes for access to non- Party gains privileged access to major
communication communication party channels of communication channels of communication
Representative Trustee Delegate Entrepreneur Agent of state
style
Source: Katz and Mair (1995)
T A B L E 6 . 4 Percentage of British MPs whose “main prior Profession” was “politician/political organizer”
1964 1966 1970 F’74 O’74 1979 1983 1987 1992 1997 2001 2005 2010 2015
Conservative 3.6 0.8 3.0 1.7 1.4 2.4 3.0 5.6 5.9 9.1 10.8 10.1 10.1 12.1
Labour 2.2 2.5 3.8 2.7 2.5 5.0 3.3 5.2 8.9 9.6 10.7 16.9 20.2 25.4
Total House of Commons 2.9 1.9 3.4 2.1 2.0 6.8 3.2 5.4 7.3 9.5 10.5 14.1 14.5 17.1
Source: Lukas Audickas, Oliver Hawkins, and Richard Cracknell, UK Election Statistics: 1918–2016, House of Commons Library Briefing paper Number CBP7529, July 7,
2016, Tables 11a and 11b

direct subventions and/or for tax advantages, such that the state is now the
largest source of funds in most European democracies (see Chapter 5; van
Biezen and Kopecký 2017).
As Bolleyer (2011) has shown, unpacking the concept of stratarchy can be
quite challenging. As employed by its originator (Eldersveld 1964), it is
primarily about mutual independence—what might in the contemporary
parlance of Europeanization be called subsidiarity or internal federalism. As
employed by Carty (2004), it is more about functionally defined division of
labor. From either perspective, however, increased stratarchy would be indi-
cated by increased autonomy or differentiation of subnational party units.
While the evidence is drawn primarily from federal (or federalizing/regional-
izing) countries, Table 3.4 and earlier work by Detterbeck and Renzsch (2003)
show an increasing tendency to subnational diversity in coalition formation.
While there is also evidence of increasing central influence over local nomin-
ations for national office (interpreted in Chapter 3 as evidence of the central
party in public office attempting to reduce the influence of activists in the
party on the ground), this only challenges the idea of stratarchy in the Carty
sense;14 in the Eldersveld sense, the real question would concern (changes in)
the level of central party intrusion into the selection of candidates for local or
regional offices, and here, at least for Italy, Wilson’s (2016) work suggests
increasing local autonomy.
As party memberships have declined, many parties, taking advantage of
possibilities offered by the internet and anxious to maintain at least the appear-
ance of grassroots connections, have experimented with forms of affiliation
short of traditional membership. As Kosiara-Pedersen and her co-authors
suggest (2017), this often implies regarding membership as a “consumption
good” for which a party can increase demand by lowering the price rather than
as an expression of commitment. Moreover, in granting privileges, sometimes
including the right to participate in candidate- or leadership-selection contests,
to these members-lite or “tourists” (see Chapter 3), the party blurs the distinc-
tions, on the one hand between member and financial contributor, and on the
other hand between member and electoral supporter.
The final three attributes listed in Table 6.3 can be reviewed quite briefly.
With regard to bases of support, as we saw in Chapters 2 and 4, the cleavages
that underlie support for mass parties have been substantially eroded. The
capacity of parties to compete on the basis of benefits provided has been
severely limited by economic and demographic realities. Instead, mainstream
parties compete on the basis of their managerial capacity and perceived
rectitude; not who will do what, but who will do it better and more honestly.
14
On the other hand, Detterbeck (2016) reports that, notwithstanding gender quotas, the
implementation of which has often been used to justify greater centralization, subnational elites
still predominate in the election of parliamentary candidates in Germany.

The institutionalization of televised debates among presidential or prime
ministerial candidates, and of party political broadcasts, have given the
leaders of parties deemed to be strong enough to warrant inclusion privileged
access to one of the most significant channels of communication. And since it
is very difficult to build the support that would qualify one for inclusion
without first having been able to use those very channels, this effectively
privileges the status quo. Finally, as politics becomes a profession in its own
right, as the boundaries between the upper reaches of the politically neutral
civil service and political appointees become more fluid, as parties depend
increasingly on state-provided resources and are governed by state-imposed
rules (notwithstanding that the legislation granting those resources and
imposing those rules was ultimately written by the parties themselves),
and as parties compete over capacity to manage policies that do not funda-
mentally change regardless of who occupies ministerial office, the boundary
between the parties and the state itself becomes more murky. In the end,
parties tend to become agents of the state, not so much in the sense that the
state acts as a principal that uses the parties as in the sense that the parties
have become a part of the state apparatus.
None of these trends has advanced to a stage at which one could point to a
particular party and identify it as being a cartel party in the full sense of the
ideal type, but neither were there ever any fully fledged mass parties or catch-
all parties. That said, however, there is reasonably strong and consistent
evidence that the mainstream parties in Western democracies—and we
emphasize that our thesis was always limited to those parties—have been
moving toward that model, and correspondingly away from the mass and
catch-all models.
THE CARTEL PARTY, CONSENSUS DEMOCRACY,

AND THE REGULATORY STATE
In the traditional party government model of democracy with which we

began Chapter 1, the voters are presented with clear alternatives, be they
advanced by individual parties or by coalitions of parties, and be they
alternative packages of policies or alternative teams of leaders. It resolves
the paradox that parties are expected to play two potentially contradictory
roles, on the one hand representing particular elements of society before the
government, and on the other hand being the government and ruling in the
interest of the entire society, by equating the will of the majority with the will
of the whole.

An alternative way to deal with the paradox is to assume the existence of a
singular national interest. This difference in orientation is reflected in
Lijphart’s (1999) distinction between majoritarian and consensus democracy.
It is also reflected in the distinction between the party government model and
contemporary models of governance or the regulatory state. And, at least if
one accepts Finer’s (1970: 8, italics in original) definition of politics as what
happens when “a given set of persons . . . require a common policy; and . . . its
members advocate, for this common status, policies that are mutually exclu-
sive,” it also suggests the possibility of a distinction between a democratic
politics of choice and an alternative more managerial version of “democracy,”
such that one could favor democracy and yet dislike competitive politics
because maintenance of the mutual exclusivity of preferred policies subverts
the singular national interest.
Lijphart defines democracy in the “the most basic and literal” way as
“government by the people or, in representative democracy, government by
the representatives of the people” (Lijphart 1999: 1). His majoritarian
model is basically two-bloc, policy-oriented party government. The repre-
sentatives are parliamentary Fraktionen (or constituency-level party dele-
gations, which because of his implicit assumption of party unity are
effectively the same thing). The form of representation is delegation—
the parties do what their supporters want. The standard for measuring
the success of representativeness of the parliament is whether the party
(or bloc) with a majority of the votes (which means that its policies are
supported by a majority of the citizens) has effective control of the parlia-
ment, and thus of the government. This is essentially the model expressed in
Figure 1.1.
Turning to Lijphart’s consensus model, we can say three important things
about representation, all of which must dilute this simple party-as-agent-
voters-as-principals sense of representation. First, if, as Lijphart says (citing
Kaiser 1997), “negotiation democracy” is to be a synonym for consensus
democracy, then the representatives must be able to negotiate, but this implies
a freedom of action far more consonant with a trustee than with a delegate
role orientation. Second, strong and independent central banks, strong and
independent constitutional courts, strong neocorporatist institutions all are
among the defining characteristics of consensus democracy; if Lijphart’s
“most basic and literal” definition of democracy is to be maintained, then
these must somehow be judged to be “representatives of the people.” Even
more than with parties negotiating in parliament, these must also assume a
trusteeship understanding of representation. Particularly with courts and
central banks, the very definition of “independence” means that they are
isolated from any direct political connection to the people. But the only way
to reconcile independence with representation is to allow acting in one’s

professional assessment of the interests of the people to count as being a
representative of the people.
The same problem arises in considering the democratic credentials of the
managerial state; to the extent that authority is delegated to experts, they must
in some way be considered to be representatives if the system is not to be
recognized to have become less democratic and more technocratic. Effect-
ively, supporters of consensus democracy and of a managerial state (see for
example Majone 1997) want to argue that since the delegation of authority is
done by legislation enacted by politicians chosen in periodic elections, the
experts can be relied upon to act as honest agents of the public. But this, in
turn, requires two assumptions. The first is that a neat distinction can be
drawn between ends (which continue to be decided indirectly by the people
through their elected representatives) and means (which are best left to
experts). If that distinction cannot be maintained, and if a professional
consensus develops among the experts that implicates ends as well as means,
then it is not clear how accountability through politicians can be enforced.
This is all the more problematic if the second assumption—that the politicians
see the delegation to experts simply as a way of increasing effectiveness but
not also as a way of shifting or evading responsibility—is violated, as indeed
we have argued it increasingly is violated.
This then leads to the third point, which concerns the standards Lijphart
wants to apply to judge consensus democracy. Although not quite a defining
characteristic, it is clear from his evaluation of consensus democracy versus
majoritarian democracy that Lijphart regards the proportional presence
of women and ethno-racial minorities as an important measure of the quality
of democratic representation, the realization of which he particularly associates
with consensus democracy (1999: 280–2). A second standard of success is
limiting inflation (265–8), unemployment (268–9), strikes (268), and violence
(270–2). A third is minimization of the difference in satisfaction with the way
democracy works between winners and losers, identified as supporters of parties
in or out of government at any particular time (286).15 All of this adds up to
“good government,” but aside from the fact that not all of these “goods” are
self-evidently good,16 the result is justification or legitimation on the basis of
outputs—that is, a technocratic or managerial standard of the good.
15
Lijphart clearly wants this datum to be interpreted positively, indicating relatively high
satisfaction notwithstanding loss on the part of the “outs.” An alternative interpretation is that
since another expectation of consensus democracy would be minimal policy change whoever wins
a particular election, both winners and losers are equally indifferent about electoral politics.
16
Inflation, for example, is bad for creditors but may be good for debtors; strikes may be the
best way to advance the interests of unionized labor; concern with representativeness with respect
to gender or ethnicity may come at the expense of concern for policy, and in any event, not
everyone agrees that these are desirable ends.

Although the political theory associated with the elite party was, as sug-
gested in Chapter 2, consistent with these consensus/managerial/unitary
national interest ideas, the democratic theory of the mass party is that of
majoritarian democracy, or the traditional model of party government: soci-
ety is divided into classes (or otherwise defined segments) that have competing
interests; each party represents the political interests of a particular class; each
party presents the electorate with a package of policy proposals that have
been developed within the party, and competes by mobilizing its clientele; the
party that wins an election then acts as the agent of its clientele to put the
policy package it promised into effect.
The replacement of the mass party by the catch-all party as the dominant
form of party organization was accompanied by a shift, or perhaps more
accurately a schizoid bifurcation, in the dominant understanding of parties
and their proper role in democratic societies. The ideas of parties as agents of
society, and as organizations that should be directed by their base members,
continued to play a significant role in political discourse and in some scholarly
analyses. At the same time, however, this principal-agent understanding of the
role of parties was undercut by the pluralist idea of parties as brokers among
various interests (Truman 1951; Dahl 1956), and by the idea that continuous
compromise to find the “middle ground” is superior to one side winning and
the other losing—even if over the long haul they alternate in office. This idea
of democratic politics gives parties a far more directive role than mere
“agency” would suggest. Indeed, as Dahl (1961, 6) says, “a leader who
knows how to use his resources to the maximum is not so much the agent of
others as others are his agents.”
The democratic theory of the cartel party is, in a sense, an exaggeration of
the pluralist (or liberal elitist) theory. Democracy is synonymous with the
capacity of voters to choose among a menu of political parties, which in turn
claim the right to occupy public office on the basis of the votes they receive.
Parties are alliances of professionals, not associations of citizens, and voters
do and should choose on the basis of “qualifications” and management skills
rather than policy proposals.
“Synonymous” is a far stronger word than “required” (as in, “elections
are required for democracy”), particularly because of what it leaves out.
The older models of democracy assumed that there would be turnover in
office, and that the possibility of such turnovers would provide a power-
ful incentive for politicians to be responsive to the desires of the citizens.
A key element of cartelization, however, is to obscure the distinction
between the “ins” and the “outs,” and by reducing the costs of being “out”
to reduce the power of the electoral incentive. And here we recall that a lack
of difference in satisfaction between winners and losers—what difference
does it make who wins?—is one of the “goods” that Lijphart claims for
consensus democracy.

In this view, democracy is transformed from a process by which the state is
controlled by civil society into a service that is provided by the government,
much as the state provides physical security through its police forces. But
since democracy, in this understanding, requires elections among political
parties, the state assumes responsibility for providing, or at least guaranteeing
the provision not only of elections, but of political parties as well.
The electoral process performs a number of functions that are valuable for
the state, beyond the possibility of coopting the public that Ginsberg (1982)
has raised. Parties and elections provide for the peaceful renewal of political/
government leadership, even if they are largely confirming decisions made
elsewhere. Even if the consequences for policy or the well-being of individual
politicians are minimal, elections provide feedback to the government regard-
ing general levels of popular satisfaction. While the form of electoral compe-
tition remains, in substance it becomes competition about managerial
competence rather than alternative policies, which in effect makes all poten-
tially governing parties mouthpieces for and defenders of the policies of the
state. As Antony Jay and Jonathan Lynn have Sir Humphrey Appleby say in
an episode of the British sitcom Yes, Prime Minister:
We run a civilized, aristocratic government machine, tempered by occa-
sional general elections. Since 1832, we have been gradually excluding the
voter from government. Now we’ve got them to a point where they just
vote once every five years for which bunch of buffoons will try to interfere
with our policies.
But put in these terms, parties collectively become agents of the state, per-
forming a variety of services for the state in exchange for financial support
(direct financial subventions both to the parties’ central offices and to the
parties in public office, provision in kind of media time, etc.). Or in more
extreme form, they effectively become part of the state, with the political
leaders joining the Sir Humphrey Applebees in trying to contain the impact of
elections.
Overall, this suggests a striking similarity between the politics of consensus
democracy and the managerial or regulatory state, on the one hand, and the
politics associated with the cartel party, on the other. In both cases, there is a
blurring of distinctions between winners and losers, a lack of popular directive
power over parties in public office, and a delegation—or shedding—of
responsibility on the part of elected representatives. In fact, both appear to
lead to models of representation that are highly truncated, and that might
perhaps be described as throwbacks to the pre-democratic era of the elite
party, in which king or parliament as a whole were assumed to be the trustee
of the nation, and while individual representatives in parliament might
express the concerns or grievances of their constituents, they did not exercise
real decision-making power. The objective of government on behalf of the

people remains, but without the effective ability of the people either to decide
for themselves what that means, or to reward and punish those who claim to
be acting as their trustees. This similarity is particularly striking, given that the
emergence of a cartel party system is generally seen as a danger to democratic
government, whereas the model of consensus democracy often is advanced as
a goal toward which democracies should be encouraged to move.
Paralleling the fact that neocorporatist consensus between the “social
partners” of business and labor is easier to maintain in times of economic
expansion, when the negotiations concern the sharing of gains, than in times
of contraction, when negotiations concern the allocation of pain (e.g. Streeck
1993: 81–2)17, the “output legitimation” (Scharpf 1999) that underlies the
consensus/managerial/cartel systems is easier to maintain when times are
good. This was particularly evident with regard to the “permissive consensus”
regarding the development of the EU: so long as the results were good, and
the major parties were all in agreement that further integration was both
desirable and necessary,18 the public were willing to let the elites get on with
it. But as Lindberg and Scheingold observed as early as 1970, if “the perceived
gains of integration were threatened . . . there [would] be reason to suspect that
the level of support or its relationship to the political process would be
significantly altered” (277), and by the mid-2010s, with the British having
voted to leave the EU, and the possibility of further exits clearly on the table,
the permissive consensus at the mass level could no longer be relied upon.
At the same time, however, and in keeping with our argument, the consensus
among the mainstream parties remained evident; for example, although the
issue clearly divided the Conservatives, the leaders of all of the major British
parties supported the “remain” position in the Brexit referendum in 2016. In
the 2017 general election, the Liberal Democrats (perhaps responding to their
electoral drubbing in 2015 and feeling a need to differentiate themselves from
their erstwhile coalition partners) and the SNP (perhaps a mainstream party
in the context of Scottish politics, but hardly in the Westminster mainstream)
continued to oppose Brexit, but the two major parties (between them,
85 percent of the seats in the House of Commons after the 2015 election
and 89 percent after 2017) both switched sides to support Brexit.
Looking beyond Britain, and beyond the question of membership of the
EU, three trends appear to be common to most established democracies.
First, whether driven by external factors (demographic change, foreign
affairs, debt crises), an increasingly shared sense of what is “responsible”
17
Other factors that Streeck suggests undermined neocorporatism include the lessened ability
of governments to reward wage restraint with other benefits and economic changes that lessened
the ability of peak organizations to speak for, or control, either labor or employers.
18
The necessity of further integration was often expressed through the metaphor of riding a
bicycle: if the European project did not constantly move forward, the result would be a crash.

(itself perhaps driven by increasing complexity and thus reliance on commu-
nities of experts with their own shared sense of what is appropriate), a shared
desire to evade responsibility (by shifting decisions to non-partisan agencies),
the policy proposals of the mainstream, or cartel, parties have come increas-
ingly to resemble one another. Second, even while the official policies of the
mainstream parties converge, dissensus with and among their supporters has
grown—reflected in declining levels of attachment on the part of their trad-
itional support groups and declining vote shares for the mainstream parties
taken as a group. In this respect, the selection of Donald Trump as the 2016
nominee of the American Republican Party or the 2015 selection of Jeremy
Corbyn as the leader of the British Labour Party, both against the overwhelm-
ing opposition of the established party elites, shows why those elites might
want to disempower not only activists, but party members/supporters more
generally. Third, the big winners have been parties outside of the traditional
mainstream.19
19
For example, although the Dutch election of March 2017 was reported as a “victory” for
Liberal Prime Minister Mark Rutte (“Dutch election result: Mark Rutte sees off Geert Wilders
challenge as Netherlands rejects far-Right” (Telegraph, March 16, 2017)), the VVD vote actually
declined by 5 percent, while the total vote of the four major parties (VVD, CDA, D’66, PvdA)
dropped by over 15 percent in comparison to 2012. Conversely, the right populist PVV gained
3 percent and the GroenLinks (Green Left) gained 6.6 percent.
The Cartel Party and Populist Opposition
In contrast to many previous models of party development that have at least

implicitly claimed to have identified an end state for party organizations, our
more dialectic approach (each new party form stimulates an adaptation by its
opposition, leading to an endless series of thesis-antithesis-synthesis) explicitly
postulated a reaction to party cartelization—which we labeled anti-party-
system parties (Katz and Mair 2009: 759) or anti-party parties (Katz and
Mair 2002: 134). When we originally advanced this idea, our primary point of
reference was the range of anti-tax and extreme nationalist parties like the
Progress Parties in Scandinavia, the Front National in France, National
Action in Switzerland, the Vlaams Bloc in Belgium, or the Freedom Party
in Austria. While we found them to be of some concern, they did not seem to
be a cause for immediate alarm. After all, in 1995, the Front National had
never won more than 15 percent of the vote in a French presidential election,
and except for 1986 (35 seats out of 573) when proportional representation
temporarily replaced the two-round majority election system for parliamen-
tary elections, they had never won more than a single seat in the National
Assembly. The best result for the Norwegian Progress Party was 13 percent in
the 1989 Storting election, but that had been more than halved in 1993; their
Danish counterpart, which had won 15.9 percent of the vote in 1973, was at
under 7 percent; the far-right Sweden Democrats had yet to win a single seat in
the Riksdag. The best showing of National Action was five seats in the Swiss
Federal Assembly (1991, fallen to three in 1995). The Vlaams Bloc had not
reached 15 percent in the Flemish Parliament, and was well under 10 percent in
the national parliament. Only in Austria had the far right achieved as much as
20 percent of the vote or taken part in government. Moreover, there was a
history of populist parties like the Poujadistes in France or Uomo Qualunque in
Italy that might have had a noticeable impact in one or two elections, but had
then disappeared. On the left, parties like the Greens might have started as
outsiders, but were in the process of joining the system. (On what at the time
seemed to be a tendency to exaggerate the system-threatening importance of
the Greens and other new parties, see Mair 1993).
Twenty-plus years on, the situation looks quite different. As we had sug-
gested might happen (Katz and Mair 1995: 24), anti-party-system parties
(some among those to which we had originally pointed and others that in

1995 seemed too trivial to mention or that had not yet been founded) have
made great gains “based on their assumed capacity to break up what they
often refer[red] to as the ‘cozy’ arrangements that exist between the estab-
lished political alternatives”—in other words, based on their desire to chal-
lenge what we have identified as a cartel. In 2016, a Green (albeit standing as
an Independent) was elected president of Austria, barely defeating the candi-
date of the FPÖ in the initial runoff election;1 in the first round, the candidates
of the SPÖ and ÖVP, the mainstream parties that had dominated Austrian
politics since the war, could only muster 22.4 percent of the vote between
them, placing fourth and fifth, respectively. For several months before the
March 2017 election, it appeared possible that the Dutch PVV (which had
already been in government briefly in 2010) would become the largest party in
parliament, and although they ultimately came a distant second, they still
gained five seats over their result in 2012 (when they had been an even more
distant third). Despite polls published only three days before the election
showing National Front candidate Marine Le Pen to have the support of 41
percent of the voters who expressed a preference (Le Monde, May 4, 2017: 2),
she ultimately received only 34 percent of the vote in the French presidential
election of 2017—but that was nearly twice the vote received by her father in
2002 (17.8 percent), the only other time a candidate of the Front National had
advanced to the second round. Moreover, in an only slightly muted echo of
the Austrian election, the candidates of the two mainstream parties (Repub-
licans and Socialists) only managed to finish third and fifth in the first round,
between them winning only a bit over 26 percent of the vote.
Table 7.1 reports the growth of several of these parties in Europe. Missing
from this list is Switzerland, in which the populist Swiss Democrats (the
former National Action) and Swiss Freedom Party were absorbed into the
Swiss People’s Party, which in the 1990s had adopted a more right-populist
agenda and in 1999 won 22.5 percent of the vote (up from 14.9 percent in 1995
and 11.9 percent in 1991), securing a second seat in the Federal Council after
the 2003 election in which they achieved 26.6 percent of the vote. And, of
course, except in countries like Switzerland (or the United States, with the
election of Donald Trump as a Republican), where the populists advanced by
effectively taking over one of the established governing parties, the rise of
populist support implies a loss of support by the erstwhile mainstream. (For
examples of the declining vote share of the old mainstream, see Table 7.5.)
The combined effects of the rise of new parties and the decline of support
for the old mainstream parties is reflected in sometimes quite substantial
1
The initial margin (May 22) was only 0.6 percent, but this result was annulled by the
Constitutional Court which ruled that postal votes had been illegally and improperly handled.
(Before the postal votes were added, the FPÖ candidate actually had been in the lead.) In the rerun
of the second round, the Green won 53.5 percent of the votes.
T A B L E 7 . 1 Recent performance of populist parties in Europe
UKIP (UK) In first place with 27% of the vote and 24 MEPs in the 2014
European election—up from 16.6% in the 2009 European election
and 3% in the 2010 general election. In the 2015 general election
UKIP fell to 13%, still far above its 2010 showing, and as in 2010 in
an election held under FPTP rather than PR, with all the biases
against small parties that FPTP entails. In the 2017 general election,
UKIP only contested 377 out of 650 seats, and won none.
MoVimento 5 Stelle (Italy) In second place in both the 2013 general election and the 2014
European election, with 25% and 21% of the vote, respectively—the
first national elections for a movement only founded in 2009.
Podemos (Spain) Winning 8% in the 2014 European elections only four months after
the party’s founding and 20.7% of the vote in the 2015
parliamentary election.
Jobbik (Hungary) Winning 14% in the European election and 20% in the national
election in 2014—roughly equivalent to its performance in the
previous elections, but up from 2% in 2006.
Partij van de Vrijheid 5.9% in its first national election (2006), rising to 15.4% in 2010 and
(Netherlands) then going to 10.1% in 2012 and 13.1% in 2017.
Alternative für 4.7% in the 2013 federal election, rising to 7.9% in the 2014
Deutschland (Germany) European election. In 2016 Land elections, the AfD won more than
14% of the vote in Berlin, more than 15% in Baden-Wüttemberg
and more than 20% in Mecklenburg-Vorpommern. In the 2017
federal election, 12.6% of the vote and 94 seats in the Bundestag,
scoring as the most popular party in Saxony and coming second in
four more Länder.
True Finns (Finland) Less than 5% of the vote in the parliamentary elections of 1999,
2003, and 2007, but then 19% in 2011 and 17.6% in 2015. After
2015, they joined a governing coalition—after which their support
as registered in opinion polls declined to under 9% by the end of the
year.
ΣΥΡΙΖΑ [Syriza] (Greece) Founded in 2004 as an alliance, winning 3.3% of the vote and then
5% in 2007 and 4.6% in 2009. Refounded as a party after winning
16.8% of the vote in May 2012, rising to 26.9% in June, and then
36.3% in January 2015 and 35.5% in September.
Danish People’s Party Founded 1995. 7.4% in 1998 rising to 21.1% in 2015.
Sweden Democrats Under 2% until 2006 (2.9%), then 5.7% in 2010 and 12.9% in 2014.
Progress Party (Norway) Under 13% until 1997 (under half that except in 1989), then peaking
at 22.9% in 2009 before falling to 16.3% (but entering government).
15.2% in 2017.
National Front (France) Founded in 1972, the FN won less than 0.5% of the vote in 1981,
soaring to 10.9% in the 1984 European election and 9.65% (and 35
seats) in the 1986 national legislative election. In 1988, the FN
remained at 9.66% of the first-round vote, but with the return to
ballotage winning only one seat. After growing in the 1990s and
early 2000s, the FN vote fell to 4.3% in 2007, rebounding in 2007
under Marine Le Pen. With 16.86% of the vote, Jean-Marie Le Pen
advanced to the second round of the 2002 presidential election (in
which he received only 16.9% of the vote). He was fourth in the first
round in 2007 (10.4%); Marine Le Pen was third in 2012 (17.9%). In
2017, she was second in the first round (20.3%), before losing in the
second round (33.9%).
La France insoumise Founded in 2016 as a left populist movement. Its candidate (and
founder) Jean-Luc Mélenchon won 19.68% of the first-round
presidential vote in 2017. The party won 11% of the first round vote
and 17 seats in the 2017 parliamentary election.*
Note: * in contrast to 13 percent of the first-round vote, but only eight seats for the National Front

decreases in the level of concentration of party systems, particularly at the
electoral level. As we saw in Chapter 6, looking at the first election after 2000,
the party systems in ten of the thirty countries observed would have qualified
as “highly concentrated” by the standard used by the US Department of
Justice to assess economic markets; looking at the last election before 2017
in the same countries, that number had fallen to four. At the legislative level,
the drop was from seventeen highly concentrated systems to fourteen, reflect-
ing in part the tendency of most electoral systems to over-reward large parties
(which might alternatively be described as their “cartel-protecting” nature).
Indeed, as shown in Table 7.2, fragmentation with regard to votes received
has increased in twenty of the thirty countries, while fragmentation with
regard to parliamentary seats has increased in nineteen (not all of which
were among the twenty). Moreover, of the ten that did not see increased
electoral fragmentation, five were among the six that had the highest levels
of fragmentation at the beginning of this period (the sixth was Belgium, which
as discussed in Chapter 6 might better be regarded as having two separate
party systems, each of which was substantially less fragmented than the
national “system” as a whole), and two more (Poland and the United States)
only saw trivial reductions.
The increasing fragmentation of party systems coupled with the growth in
support for populist parties and candidates was accompanied by renewed talk
of a crisis of democracy or a crisis of party government (e.g., Beaumont 2011;
Mac Con Uladh 2013; Sen 2012), but it was not the only reason for concern.2
Among the symptoms contributing to the diagnosis of crisis have been
marked declines in popular involvement with, attachment to, and trust in
political parties (Ignazi 2014). In particular, there have been nearly universal
drops in both the proportions of electorates who report a party identification
and in the reported strength of party identification among those who still
identify at all (Dalton and Wattenberg 2000);3 in the proportions of elector-
ates who become formal members of parties (detailed in Chapter 3); and in
voter turnout in national elections. For example, referring to turnout in the
2001 British general election—then the lowest by more than 10 percent since
1945—Whiteley et al. (2001: 786) observed that “[t]he word crisis is often
abused in contemporary accounts of politics. But if this is not a crisis of
democratic politics in Britain, then it is hard to know what would be.” And
2
For references to earlier “crises of democracy,” see Chapter 2.
3
More recently, Schmitt (2014: 81), although observing that the decline is not uniform or fully
universal (perhaps more accurately, not universally monotonic), reports “compelling evidence of
a decline in partisanship.” Based on the European Social Survey, Hooghe and Kern (2015: 948)
show a small decline in the proportion of respondents reporting that they feel “very” or “quite”
close to a specific party in the sixteen countries that participated in all five ESS waves, 2002–10,
but even if this slight decline is interpreted as a leveling off, it is at a post-decline level.
The Cartel Party and Populist Opposition 155

T A B L E 7 . 2 Effective numbers of parties in the first election after 2000 and
the last elections before 2017
Country Year Effective number of Effective number of

electoral parties parliamentary parties
Year 1 Year 2 Year 1 Year 2 Year 1 Year 2
Australia 2001 2016 3.44 4.47 2.49 3.07

Austria 2002 2013 3.02 5.15 2.88 4.59
Belgium 2003 2014 8.84 9.62 7.03 7.82
Canada 2004 2015 3.78 3.33 3.03 2.50
Czech Republic 2002 2013 4.82 7.61 3.67 6.12
Denmark 2001 2015 4.69 5.86 4.48 5.57
Estonia 2003 2015 5.42 5.14 4.67 4.72
Finland 2003 2015 5.65 6.57 4.93 5.84
France 2002 2012 5.22 5.27 2.26 2.83
Germany 2002 2013 4.09 4.81 3.38 3.51
Greece 2004 2015(S) 2.66 4.51 2.19 3.24
Hungary 2002 2014 2.94 3.22 2.21 2.01
Iceland 2003 2016 3.94 6.08 3.71 5.09
Ireland 2002 2016 4.13 6.57 3.38 4.93
Italy 2001 2013 6.32 5.33 5.30 3.47
Latvia 2002 2014 6.78 5.60 5.02 5.13
Lithuania 2004 2016 5.78 6.79 5.46 4.42
Luxembourg 2004 2013 4.26 4.85 3.81 3.93
Netherlands 2003 2012 4.99 5.94 4.74 5.70
New Zealand 2002 2014 4.17 3.27 3.76 2.96
Norway 2001 2013 6.18 4.87 5.35 4.39
Poland 2001 2015 4.50 4.45 3.60 2.75
Portugal 2002 2015 3.03 3.59 2.50 2.86
Slovakia 2002 2016 8.87 7.31 6.12 5.67
Slovenia 2004 2014 6.02 5.33 4.90 3.97
Spain 2004 2016 3.00 5.03 2.53 4.16
Sweden 2002 2014 4.51 5.41 4.23 4.99
Switzerland 2003 2015 5.44 5.83 5.01 4.92
UK 2001 2015 3.33 3.92 2.17 2.53
USA 2002 2014 2.15 2.14 2.00 1.96
Source: Michael Gallagher, 2017. Election indices dataset at http://www.tcd.ie/Political_Science/staff/michael_
gallagher/ElSystems/index.php, accessed 24 March 2017
although British turnout increased in 2005 and again in 2010, 2015, and 2017,
even the 2017 turnout at about 68.7 percent was more than 2 percent below
the 1945–97 low.4 Similarly, Italian turnout of just over 75 percent in 2013
was more than 5 percent below the previous post-war low (International
4
Turnout figures are from UK Political Info: http://www.ukpolitical.info/Turnout45.htm.

IDEA, Voter Turnout Database).5 In other words, at least two of the trends—
declining turnout and declining major party vote share—that Crewe (1974: 82)
identified as “peculiar to Britain” have become the norm for the industrialized
democracies.
The natural question is what has produced this decline in support for the
established parties, and the concomitant rise in support for anti-party-system
parties—in particular those that are fairly identified as “populists.” A focus on
the rhetoric and recent successes of the parties cited in Table 7.1 suggests that
the explanation should be sought in similarly contemporary events, and there
is no shortage of proposed explanations for the presumed crisis. Many recent
invocations of a crisis of party democracy in Europe have attributed the
problems to failures in the handling of the 2008 global financial crisis, the
slightly more recent euro crisis, or the even more recent refugee/immigration
crisis (Sen 2012; Bellamy and Staiger 2013; Collignon 2015; Beaumont 2011;
Mac Con Uladh 2013). But if, as we observed in Chapter 2 and as is, for
example, reflected in the titles of numerous works from the 1970s, The Crisis
of Democracy (Crozier et al. 1975),6 or a series of crises of democracy, has
been going on for decades, then it may be important to consider more general
and long-standing causes.
Cas Mudde (2010: 1175) defines populism as “a thin-centred ideology that
considers society to be ultimately separated into two homogeneous and antag-
onistic groups, ‘the pure people’ versus ‘the corrupt elite.’” Thus one of the
recurring populist claims, on both left and right, is that “corrupt elites have
hijacked the political system and silenced the voice of the people by making
backroom deals and enforcing a conspiracy of silence” (Mudde 2013: 7, albeit
writing only about the “populist radical right”). As we suggested in our 1995
article, this is what the anti-party-system parties generally have done—accuse
the mainstream parties of feathering their own nests (and incidentally the nests
of their elite “co-conspirators”) while ignoring the concerns of ordinary citi-
zens. Both this and the constrained competition typical of cartelization were
highlighted by the performance of Philippe Poutou, the candidate of the New
Anticapitalist Party, in the French presidential election debate of April 4, 2017.
Poutou took less than a minute “to puncture the mutually protective world of
5
Turnout in Austria in 2013 (74.91%) was a new post-war low, as were the French turnout in
2012 (55.4%) and Portugese turnout in 2011 (58.03%) and then 2015 (55.2%). Belgian turnout
in 2014 (89.22%) was only 0.15% above the post-war low set in the previous election, while German
turnout in 2013 was less than 1% above the post-war low set in 2010; while it rebounded a bit in
2017 (to about 76%), that was at least 10% below what had been the norm through the early 1980s.
Turnout in the 2017 Dutch election was the highest since 1986, driven partially by unusually good
weather on polling day, but also by the sharp contrast between mainstream Prime Minister Rutte
and Geert Wilders; nonetheless, it too was substantially below the pre-1980 norm.
6
Other examples include Must Canada Fail? (Simeon 1977), and “Italy—Ungovernable
Republic?” (Allum 1974).

the race’s mainstream” and to show “his ‘lack of respect,’” highlighting the
scandals plaguing the front-running candidates—scandals that, although
covered in the print media, had “been mostly taboo beneath a well-maintained
facade of respectability” in the debates (Nossiter 2017).
No doubt, this perception has been re-enforced as the direct result of actual
malfeasance on the part of party politicians. When a significant percentage of
the national economy is being paid as bribes and kick-backs to the parties, as
was the case in the Italian tangentopoli scandal, or when members of the
British parliament are found to have been using parliamentary expense claims
to pay for the cleaning of country house moats or the like and at least one of
the MPs implicated ends up with a seat in the House of Lords (Stone 2015), or
when a former Greek minister is only given a suspended sentence after
admitting that he took over €200,000 in bribes (Apokoronasnews 2015), it is
easy to understand why citizens would become cynical about parties and party
politicians. Even without overt illegality, the Italian practice of lottizzazione
(and the equivalent system of Proporz in Austria, or elements of consoci-
ationalism in the Netherlands) would comport well with the idea that the
parties have long seen the public sector as a pot of spoils to be plundered for
their own advantage.
We have already referred to lottizzazione, Proporz, and the sharing of
public-sector jobs and public subsidies among the consociational partners in
the Netherlands as indicators of cartelization, although we identified the
strategy of cartelization as largely defensive (a reaction to threatened loss of
position/privilege) in contrast to a more populist characterization of these
actions as aggressive (a conscious rent-seeking strategy driven by aggrandiz-
ing self-interest). From the perspective of “the people,” however, this is in
many ways a distinction without a difference. Popular demands go unheeded,
and indeed unexpressed by the parties in power or perceived to be seriously
challenging for power. Policies appear to favor powerful interests rather than
ordinary people—and to change only at the margins, if even there, regardless
of which particular parties are in government. Exemplifying this with regard
to domestic policy, although a 2015 Institute for Fiscal Studies analysis of the
post-election plans of the major British parties found differences between the
Conservatives and Labour that they described as “substantial,” the real
conclusion was that “All four parties’ plans imply further austerity over the
next parliament” (Election 2015: Briefing Note 11, p. 2). At least with regard
to the mainstream parties, interparty competition has been hollowed out—the
form remains, but with decreasing substance.
In this chapter, we pursue the idea that although current issues like immi-
gration or economic dislocation may be the catalysts for the current wave of
populist revolt, the underlying causes are inherent in the contemporary prac-
tice of democratic party politics, and thus in the processes that have driven
cartelization itself.

TIES BETWEEN CITIZENS AND PARTIES
As shown in Chapter 4, one element of cartelization has been the constriction

of what might be called the “active partisan policy space”—that subset of the
range of issues that might be subjects for policy debate that actually are
addressed by mainstream parties, and the range of positions actually taken
by the mainstream parties on those issues that they do address. One way in
which this has been manifested has been a downplaying of many of the
sociological divisions that undergirded the identity politics of the mass
party, and indeed were the foundation of the party-defining cleavages that
Lipset and Rokkan (1967) identified as the bases of European party systems
from the 1920s into the 1960s.
Perhaps the most obvious of these is the religious cleavage between Prot-
estants and Catholics or between Catholics and seculars. Examples regarding
the reduced significance of the former include the German Christian Demo-
cratic Union, founded in 1945 as an interconfessional (Catholic plus Protest-
ant) party in reaction to the way in which antagonism among the heavily
Catholic Zentrum and the heavily Protestant German Democratic Party and
German People’s Party facilitated the rise of the Nazis, and the Dutch
Christen-Democratisch Appèl, founded as an alliance in 1973 and as a new
party in 1980 through the merger of the Katholieke Volkspartij with the
Protestant Anti-Revolutionaire Partij and Christelijk-Historische Unie.
With regard to the division between Catholics and seculars, perhaps the
most obvious example is the Belgian Parti Social Chrétien, which in 2002
changed its name to Centre démocrate humaniste. Similarly, when the
Christen-Democratisch Appèl formed a coalition with the liberal (secular)
People’s Party for Freedom and Democracy, it became more difficult to
maintain the importance of religion as a fundamental political cleavage in
Dutch politics. Less directly (because it did not specifically refer to Christian
symbols, and was widely interpreted to be directed at Muslims), when the
French Code of Education was amended in 2004 to ban the wearing of
conspicuous religious symbols in state-operated schools, the legislation
was supported by all of the major parties (UMP, UDF, PS—collectively
over 520 seats in the 577-member National Assembly), although there was
division within each of them.
Beyond these internal divisions, as repeated controversies over the placing
of crucifixes in Bavarian classrooms show,7 the question of religion in educa-
tion (often the focal point of historical religion-secular disputes) has not
7
See, for example, “Crucifix Ruling Angers Bavarians,” New York Times, August 23, 1995;
“Father sparks classroom crucifix row in Bavaria”, November 17, 2010, https://www.thelocal.de/
20101117/31229.

entirely disappeared from politics, particularly in heavily Catholic areas.
At the same time, however, the Bavarian controversies differed from those
historical disputes in two important respects. First, the question was about the
display of Christian symbols, and not about control over the content of
education, which in accordance with Article 7 of the Basic Law remained
under the supervision of the state. Second, although the parties made state-
ments concerning the issue, it was not put on the agenda by the parties but
rather by the Constitutional Court in ruling on a case brought by a parent.8
In other words, even if these questions remain salient for some elements of
the electorate, they are not strongly reflected in differing positions of the
major parties.
One reason for the decline of religion as a point of party contention has
obviously been the decline of religious belief and practice on the part of
citizens. Another is the deliberate choice of Christian Democratic parties to
move in the direction of “catch-all-icism” and to appeal to economically and
socially conservative, but not religious, voters. (See, for example, Kalyvas and
van Kersbergen 2010.) Few people would regret the decline of interconfes-
sional animosities, but the weakening of the connection between party and
church, coupled with a weakening of ties between citizens and churches, also
weaken one of the ties of identity that kept the citizens connected to the
mainstream party system.
In response to the weakened confessional ties of the mainstream parties, a
number of new Christian parties have been formed outside of the mainstream
Christian Democratic tradition, including the Danish Kristendemokraterne
and the Dutch ChristenUnie; in other cases, such as the Dutch Staatkundig
Gereformeerde Partij, already existing small orthodox parties have resisted the
temptation to dilute their theological commitments in order to attract a wider
voter base. While their supporters may remain well connected both to their
church and to their party, they have deliberately placed themselves outside of
the political mainstream.9
The prominence of issues identified by the Manifesto Research Group as
belonging to the “economic cluster” has declined, as has the range taken by
mainstream parties on the resulting left/right issue dimension (see Tables 4.1
and 4.3–4.5). And as with the weakening of ties between parties and organized
8
A similar complaint was addressed by the European Court of Human Rights in the case of
Lautsi and Others v Italy. The court allowed the crucifixes to remain, based in part on a finding
that their mere presence did not amount to indoctrination and was not associated with
compulsory teaching about Christianity.
9
On the other hand, while divisions among the Christian sects, and even between Christians
and the non-religious, may have weakened in intensity, this does not necessarily extend to “alien”
religions like Islam, whose adherents may therefore replace members of alternative Christian sects
as the “other” against which identity may be constructed.

T A B L E 7 . 3 Union density, 1982 and 2013
1982 2013
Austria 54.5 27.8

Belgium 49.4 55.1
Denmark 80.2 66.8
Finland 68.4 69.0
France 17.0 7.7
Germany 35.0 18.1
Ireland 56.1 29.6
Italy 46.7 37.3
Netherlands 32.3 17.8
Norway 58.1 52.1
Spain 12.2 16.9
Sweden 78.9 67.7
Switzerland 26.2 16.2
Source: https://stats.oecd.org/Index.aspx?DataSetCode=UN_DEN#
religions, this issue convergence has been accompanied by reduced organiza-

tional connections. On the one hand, some social democratic parties have
either severed or weakened their organizational ties to unions (Allern and
Verge 2017, who note that these formal ties were never terribly common, but
have declined even further since the 1960s), while on the other hand, as shown
in Table 7.3, the density of union membership has generally declined, and
where union density has remained high, it is often because of the unionization
of white-collar civil servants, who hardly form a reliable constituency for
traditional class-based politics. As with declining participation in organized
religion, and beyond its impact on the policy proposals on offer, the decline in
union membership also weakens one of the social bases for ties between
citizens and the party-political mainstream.
Finally, although not directly indicating political connectedness, it is likely
that the act of moving house, with the associated disruption of ties to neigh-
bors and neighborhood, would also have the potential to disrupt the mover’s
political habits. And, as Table 7.4 shows for a selection of countries, rates of
residential mobility increased quite significantly in the thirty years between
1981 and 2011.
While these changes weaken the triangular connection between citizens
and parties illustrated in Figure 7.1 by undermining the connections between
citizen and intermediating groups, especially church and unions (“A”), and
between those groups and parties (“B”), the direct connection between citi-
zens and parties (“C”) has been weakened, in two ways. At the aggregate
level, there simply are fewer party members and fewer citizens with strong
partisan attachments. Some parties, as noted in Chapter 3, have experimented
with strategies to counter membership decline, and increase the connection

T A B L E 7 . 4 Percentage of the population that changed usual
residence in the preceding year
1981 2011
Great Britain 9.6 13.1

Sweden 9.5 13.3
France 9.4 12.8
Netherlands 7.7 9.8
Belgium 7.3 11.5
Ireland 6.1 9.8
Source: 2011 data computed from http://ec.europa.eu/eurostat/web/
population-and-housing-census/census-data/2011-census; 1981 data
from Long 1991
Citizens
A C
Intermediating Parties
groups
B
F I G U R E 7 . 1 Ties to groups and parties
between the party and its members/supporters (Scarrow 2015). Regardless of

whether these strategies are successful in numerical terms (and once the
definition of “membership” is expanded to include “registered supporter,”
to use the British Labour Party’s term for a form of “membership-lite,” or
“followers” to use the Twitter or Facebook terms, it is hard to make mean-
ingful before and after comparisons), however, the fundamental nature of the
connection between party and member or supporter has changed. Rather than
attending local party meetings—and developing face-to-face relationships
with other party members—communication between members and their
party increasingly takes place over the internet and by means of bulk emails,
chat rooms, and electronic fund transfers. Even if direct member ballots
appear to increase the influence of members as a category—and as we argued
in Chapter 3, the result may actually be to lessen their influence by limiting
their choice to options approved or framed by the party center—they are

unlikely to increase the sense of personal connection felt by the individual
member.
While these trends do not lead directly to populism, they leave citizens less
anchored in the political sphere, and as Bernard Berelson suggested in the
final chapter of Voting (Berelson et al. 1954)—a chapter clearly inspired at
least in part by the fascist experience in Europe, from which his co-author
Paul Lazarsfeld was a refugee—this lack of connection to “responsible”
parties, or to friends or neighbors or colleagues who are anchored to such
parties, leaves citizens more vulnerable to populist/demagogic appeals should
a perceived crisis arouse them from the inactivity that would normally be
expected to characterize the disconnected.
EXPERTISE AND OUTPUT LEGITIMATION
All governing ultimately involves questions of preference or taste (what is

the objective?) and questions of technique (how can the objective best be
achieved?). In the naive principal-agent models of party government repre-
sented by Figures 1.1–1.3, electoral competition among parties allows the
electorate to choose which politicians, acting as agent, will decide the ques-
tions of preference, while technical experts in, or working with, the bureau-
cracy as the agent of the ministry dutifully answer the questions of technique.
But because the distinction between means and ends is never as clear in reality
as it appears in rhetoric, manipulation of the distinction between preference
and technique becomes a political tool. This is most commonly understood as
a strategy of bureaucrats, or other individuals whose influence rests on sup-
posed technical expertise, to increase that influence: by framing decisions as
questions of technique rather than questions of preference, they are moved out
of the political arena, and the expert is given a privileged position to decide
them, while preferences expressed from the public can be defined as naive,
ill-informed prejudices that are inappropriate to the complex and technical
question at hand.10 This less cooperative view gives rise to the portrayal in the
British comedy Yes, Prime Minister of the clash between “the political will
and the bureaucratic won’t.” In this scenario, the politicians are expected to
10
In extreme form, this implies an incompatibility of technocracy and democratic party
government analogous to the incompatibility with populism—both would deny the essential
position of party competition, in the case of populism because the common good is not to be
found through popular competition among partial interests, and in the case of technocracy
because popular preferences (aside from the assumed preference to be governed “well” or
“effectively” or “efficiently”) are irrelevant to identification of the common good.

emphasize instead the political nature of the question, thereby moving it back
into the arena in which the people, and the parties as their agents, are entitled
to decide. It must be remembered, however, that this assumes that politicians
are primarily motivated by a desire for the power to decide, rather than by
more personal motivations like job security and career advancement, an
assumption undermined by the increasing professionalization of politics.11
In many respects, the focus of democratic politics over the last several
decades has shifted from the advancement of programs or interests (questions
of taste) to problems of management (questions of technique). First, there has
been tremendous growth in the responsibilities of government. On the one
hand, governments regulate a far wider range of activities, and in a far deeper
sense; while earlier governments may have regulated banking and the money
supply, set tariffs and other trade policies, they did not assume responsibility
for managing the economy as a whole in the way that governments did from
the mid-twentieth century.12 On the other hand, governments became the
primary suppliers of a range of goods and services, including education,
transportation, health care, and pensions. While these have added to the
range of questions on which governments might be expected to make policy
choices for which “taste” would be relevant, in reality governments come into
office with most of these decisions already made by their predecessors, with
the new government left primarily to manage the status quo (e.g., Rose and
Davies 1994). One reason why government is seen increasingly in managerial
terms is that all of those increased activities mean that there is a lot more to
be managed.
A second reason for this shift of emphasis to questions of management is
that the question of whether or not the state would expand in these ways
largely ceased to be a focus of interparty conflict. Thirty years ago, one might
with some apparent justice have said that the “big” policy questions—the
welfare state; public ownership of such “natural monopolies” as transport and
public utilities; relations between church and state—all were settled, leaving
management as the primary concern by default. While public ownership has
11
In the Downsian model, politicians are presumed to use policy promises as a means to
attract votes in order to pursue their private, careerist objectives. On the one hand, this model
makes quite unrealistic assumptions about the voters’ decision calculus, while on the other hand,
to the extent that those unrealistic assumptions are modified to take account of retrospective
evaluations of performance, the model ignores the possibility that for politicians the desire to
avoid responsibility will be stronger than the desire to claim credit.
12
Although the explicit assumption of government responsibility for economic management
may be a relatively recent phenomenon, it should be remembered that even in the early nineteenth
century Henry Brougham could write “don’t forget that a Gov’t is not supported a hundreth part
so much by the constant, uniform, quiet prosperity of the country, as by these damned spurts
which Pitt used to have just in the nick of time” (Letter to Mr. Creevey, 1814 quoted in Butler and
Stokes 1969: 389).

been dramatically rolled back, and the certainty that the post-war consensus
about the welfare state was going to stay fixed now seems misplaced, and
much of it is being rolled back as well, what is significant is that the argument
has been largely couched in explicitly managerial terms, and there was
remarkably little dissent from the left to the Reaganite/Thatcherite/
Kohl-esque reforms; indeed both the substance and the rhetoric of these
retrenchments were basically embraced by social democrats Tony Blair and
Gerhard Schroeder in their 1999 statement, “Europe: The Third Way/Die Neue
Mitte.” A few years earlier, when the conservative Bildt government in Sweden
declared that welfare state retrenchment was “den enda vägens politik” (“the
only way policy”), the Social Democrats supported tax cuts and devaluation
(Agius 2007; Blyth and Katz 2005).
Earlier we suggested a number of reasons why governing parties and their
leaders of all (moderate) political persuasions would prefer to cast their jobs as
management rather than policy advancement. One is the erosion of the social
distinctions—particularly class and Christian denomination—that formed the
basis of the mass party’s explicit agenda of promoting segmental interests.
A second incentive for parties to emphasize management is their growing
recognition of the limited room for maneuver resulting from the financial
stresses brought about by the success of the welfare state and the globalization
of economic activity. At the same time, there was an ideological shift toward
neoliberalism, and the idea that the state should not even try to control the
economy.
The resulting shift from advocacy of policies to advance one or another
conflicting interest to management of a highly constrained state has been
reflected not only in a convergence of party policies, but also in changes in
the organization of the parties themselves. As we argued earlier, if party
leaders see themselves to be primarily responsible for relatively passive man-
agement rather than active reform, and are effectively bound by pragmatic
concerns, they need to free themselves from the constraints of party activists
whose primary “function” (if one may use such a word) is the making of
sectional demands. Conversely, as they increasingly justify their existence
with claims of managerial competence, the party in public office needs to
supply itself with technical expertise rather than ideological certainty.
Observations like these regarding the dichotomy between taste and tech-
nique traditionally have focused on the relations between parties/politicians
and experts (bureaucrats or technocrats) at least potentially outside of the
parties themselves. In fact, the conflict between preference and expertise is not
limited to these relations, but can be reproduced within individual parties
with regard to questions of candidate selection, policy formulation, and
campaign tactics. In this case, there are three related problems. The first is
determining what counts as expertise and where it resides. In the modern
context in which face-to-face campaigning has largely been supplanted by

mass media-based campaigning, and public opinion polling has become the
principal way of knowing what the electorate is thinking, have pollsters and
media consultants—and the fundraisers required to pay for them—assumed
a dominant role as the political experts? And conversely, has the political
expertise (“savvy”) of party members correspondingly been devalued?
The second problem in effect reproduces the tension between politicians
and technocrats in the formulation of policy, but this time concerning the role
of party members as opposed to experts chosen by (and beholden to) the party
leadership in public office and the party central office. The danger is that the
proposals of the members will be overruled by the technicians (possibly quite
properly) on grounds of practicality or realism, but that the members will
interpret this as party oligarchs running rough shod over them while the
experts act as if their professional judgment should be adequate justification.
The third problem, particularly for the party leadership in public office, is
whether the preferences expressed by members will be sufficiently strategically
informed that the members are capable of advancing their own preferences, or
whether alternatively the members are likely to be their own worst enemy.
Will they push for party decisions that are closest to the outcomes regarding
personnel and policy that they individually prefer (“if I were king”) as if the
intraparty decision would more or less automatically become the final system-
level decision? Or, will they push for party decisions that have the greatest
likelihood of resulting in a government and government policies that they
prefer—after the intraparty decision has competed and/or been reconciled
with the intraparty decisions of other parties in the electoral and parliamentary
arenas? Expressed alternatively, will party members see party primarily as a
vehicle for expression (which may include expression of ill-thought-out general
anger as well as expression of real preferences) or as a vehicle for government
entry and influence? Both of these questions point to a likely conflict between
party leaders, who hire and fire the professional experts and who enjoy the
personal rewards of electoral success, and the party’s members.13
The privileging of experts, and reliance on retrospective evaluations of
performance as a primary basis for electoral competition, have a number of
consequences for the relationship between parties and citizens. On the one
hand, to the extent that electoral judgments are based on results, there is no
need to explain and justify, or even to discuss, policies with the electorate, and
in any case, to ignore the advice of the experts raises the obvious dangers of
increasing the likelihood that things will go wrong and of having sacrificed the
ability to use the expert advice as an excuse if things do go wrong. On the
other hand, even when they disagree about specifics, experts in any particular
13
This “expertise gap” is one of the primary bases for Michels’ “iron law of oligarchy”
(Michels 1962 [1911]: 107).

field will tend to form an epistemic community (Haas 1992) with shared
backgrounds, shared values, and shared standards of evaluation—which do
not change just because the parties nominally in power change. This raises
again the dichotomy between taste and technique, because ultimately profes-
sional standards are rooted in ideologies that can themselves be characterized
as matters of taste—but the taste of professionals rather than of the people.
Two examples from the field of economic policy are the tendency of econo-
mists to take growth in GDP (or in per capita GDP) as the primary indicator
of economic success, rather than, for example, median family income, thus
privileging the “size of the pie” over its distribution, or to regard inflation
(over a certain, but not necessarily well-specified level) as naturally bad, thus
privileging the interests of those with money over the interests of those with
debts.14 In this context, it makes little political difference whether the techno-
crats are in some objective sense correct; what counts is that, on the one hand,
citizens have little independent capacity to evaluate the quality of expert
advice except by judging outcomes, and on the other hand that all of the
mainstream parties appear to accept the same expert advice, and therefore to
support what are fundamentally the same policies.
Although it is impossible entirely to disentangle retrospective judgment
based on results (reflected, among other things, in models of electoral choice
based on macroeconomic indicators like inflation and unemployment rates—
e.g., Lewis-Beck 1990) from judgment based on policy preferences, the general
phenomenon of “output legitimation” referenced in Chapter 6 particularly in
regard to the EU has hardly been limited to that question. Ronald Reagan’s
famous suggestion from the 1980 US presidential campaign that voters ask
themselves “Are you better off than you were four years ago? Is it easier for
you to go and buy things in the stores than it was four years ago? Is there more
or less unemployment in the country than there was four years ago?” is
illustrative of interparty competition based on appeals to results—and of the
underlying assumption that there is no disagreement regarding either the ends
or the policies to be pursued.
In more general terms, most Western democratic governments, and the
mainstream parties that have dominated them, have benefitted from high
levels of diffuse support. Although the concept was originally intended to
refer to support for regimes (for example, support for “party government”), it
can equally refer to the major actors within such a regime, such as the
mainstream parties as a group. In part, this diffuse support was the result of
childhood and continuing adult socialization (Easton 1975: 445), manifested
directly through party identification and indirectly through connections to
14
These are, of course, not limited to economists. See for example Arend Lijphart’s (1999) use
of low inflation and high GDP growth as unproblematic indicators of good macroeconomic
performance in assessing the relative efficacy of majoritarian and consensus democracy.

organizations like churches and unions that are related to parties. As we have
seen, both of these connections have been weakening over time. Beyond this,
diffuse support has also been maintained by decades of apparently successful
governance and economic management.15 This kind of support is always at
risk of collapse in the face of economic or other problems, however, and
indeed that risk may be heightened by too much prolonged success, which
raises expectations that in turn make even modest setbacks appear to be
serious failures; in this respect, the extraordinary sustained improvement in
standards of living in many countries after World War II may have generated
expectations of continued growth that simply cannot be realized in the ordin-
ary economy, when “productivity, economic growth, and living standards
improved haltingly, and sometimes not at all” (Levinson 2016: 9), and thereby
(pace George Will (2015)) redefined economic failure.16 Moreover, it is short-
term failures that diffuse support is supposed to cushion, so they become more
significant as diffuse support erodes. In this context, the 2015 Eurobarometer
finding that parties were the least trusted political institution in twenty-three
of the twenty-eight member states, the 2014 GfK Verein study including
eleven European democracies (fourteen if Australia, the United States, and
Canada are counted as “honorary Europeans”) that found politicians to be
the least trusted of thirty-two professions in every one except Sweden (where
retail sellers ranked even lower),17 and the 2016 Transparency International
finding that over a quarter of EU+ citizens (the EU plus Greenland and
Switzerland) report believing that their representatives are “mostly or entirely
corrupt”18 become particularly significant. At the level of individual parties,
poor results may stimulate turnover in office, or in a more extreme case the
collapse of a party, but when all of the responsible parties appear to be
committed to the same unsatisfactory positions, the result may be not a crisis
for the parties in office at the moment, but rather what might be described as a
crisis of the party system or even as a crisis of party democracy. As Jonathan
15
As David Easton (1975: 446) says, although childhood socialization plays a large role in the
development of diffuse support (cf. party identification), it also “may be a product of spill-over
effects from evaluations of a series of outputs and of performance over a long period of time.”
16
Will was accusing President Obama of trying to lower the level of performance required to
avoid being accused of failure.
17
The European democracies were: Austria, Belgium, France, Germany, Italy, the Netherlands,
Poland, Spain, Sweden, Switzerland, and the UK. Politicians also ranked at the bottom in all eleven
of the other countries included except Indonesia, where they were just “beaten” by insurance agents.
http://www.gfk.com/Documents/Press-Releases/2014/GfK_Trust%20in%20Professions_e.pdf,
accessed April 24, 2015.
18
In 2013, Transparency International found citizens to believe that political parties were the
most corrupt of twelve major institutions (in some cases, tied with other institutions) in Australia,
Belgium, Canada, Czech Republic, Estonia, Finland, France, Germany, Greece, Hungary,
Israel, Italy, Luxembourg, New Zealand, Norway, Portugal, Romania, Serbia, Slovenia,
Spain, Switzerland, UK, and United States.

Freedland (2016) observed (writing primarily about right-wing populist
appeals in the context of the 2016 American presidential election, but refer-
ring as well to “populist appeals from the left, from Bernie Sanders to
Podemos, from Jeremy Corbyn to Syriza”), “The new populists don’t simply
say that the ruling party has failed and now the opposition should have a turn.
They insist that the entire system is broken.”
WHAT IS THE PROBLEM WITH PARTY

DEMOCRACY ANYWAY?
Political crises generally result from an inability to respond to demands. In

some cases, the problem stems from limited capacity: the inability of the
Sudanese government to assure food supplies in the face of drought induced
famine; the inability of the Weimar government in Germany to satisfy the
demands of the Versailles Treaty; the inability of the Hungarian state to cope
with the global economic crisis of the late 1920s and early 1930s. In these cases,
the turn to an authoritarian government in the hope of increasing capacity is
understandable, even if it is unlikely to solve the problem. In other cases,
however, the problem results from inconsistency in the demands themselves.
For example, the 1975 Trilateral Commission report attributed what it iden-
tified as “the crisis of democracy” in the United States to the “impulse . . . to
make government less powerful and more active, to increase its functions, and
to decrease its authority” (Crozier et al. 1975: 64). The current crisis of dem-
ocracy is at least in part of this second type, spurred by the inability of current
party governments to satisfy demands, or to meet expectations, that are
themselves internally inconsistent—on at least three dimensions.
The first of these stems from the distinction just discussed between ques-
tions of taste or preference, which in a democracy implies privileging the
opinions of citizens as equals, and questions of technique, which even in a
democracy implies privileging the opinions of experts, and in particular it
stems from the fact that the idea of democratic party government ultimately
requires parties to reconcile potential conflicts between the two. On the one
hand, parties are presumed to remain the people’s representatives, organizing
and channeling demands on the state, at least some of the time being more
concerned with personal or group interests than with aggregate benefits and
costs. On the other hand, however, the parties are the government. While
politicians may not be expected to be experts with regard to the construction
of bridges or the curing of disease—or even the control of unemployment—
they are expected to be experts in the management of government. Because
the ministers, which is to say the politicians, are supposed to be in control of

the bureaucracy, bureaucratic obstruction is not an acceptable excuse for
failure—it merely identifies a different type of failure. Moreover, if “politics
is the art of the possible,” then the politicians might be expected to be experts
with regard to possibility—and not to promise the impossible, or to use
impossibility as an excuse for failure to deliver on their promises. At least
within the constraints of reality (and assuming that there is some consensus
regarding what those constraints are), the idea of responsible party govern-
ment does not allow politicians in government to evade responsibility for
results—which are often identified as the current “state of the world,” whether
or not that state of the world is plausibly the result of government actions.
But, of course, with the ceding of authority to non-partisan agents like the
International Monetary Fund (IMF) that is exactly what parties have
attempted to do.
The parties confront incompatible demands, which becomes especially
apparent in hard times. They are captives of their “experts”—although it is
important to remember that those experts generally have been chosen by the
parties themselves, often in anticipation of the advice they can be expected to
give—and of externally imposed circumstances, both of which limit their
ability to be responsive to popular preferences and to satisfy popular
demands. At the same time, they are expected to be channels for the expres-
sion of popular discontent with the resulting outcomes. And the effectiveness
of the strategy of asking the people to accept the realities of what parties can
do is limited by the history of parties claiming credit when times were better.
If parties wanted credit for good outcomes, why should they not be assigned
responsibility for outcomes that are less attractive?
Even more broadly, parties are confronted by two incompatible concep-
tions of society and democracy: on the one hand, a communitarian single-
national-interest model and, on the other hand, a liberal, individualistic,
legitimately conflicting-interests model. The first sees society as an organic
whole, with a single paramount national interest that ought to guide public
policy. While there may (probably will) be disagreements about the nature of
that interest, and those differences may imply different sets of “winners” and
“losers” with regard to particular policies, there remains a belief that all
citizens would support the same policies if only they were sufficiently enlight-
ened. Edmund Burke’s definition of a party (quoted in Chapter 2) as “a body
of men united, for promoting by their joint endeavours, the national interest,
upon some particular principle in which they are all agreed” reflects this ideal.
Essentially the same perspective was articulated nearly two centuries later by
Winston Churchill (quoted by Samuel Beer [1969: 99] in elaborating the
model of “Tory democracy”):
All I will promise to the British electorate in your name, and the only
pledge that I will give on behalf of the Conservative party is that if the

government of Britain is entrusted to us at this crisis in her fate, we will do
our best for all, without fear or favour, without class or party bias,
without rancour or spite.
Here the assumption that there is a national interest that overrides immediate
private preferences or interests such that it is possible to do one’s “best for
all . . . without class or party bias” is even more explicit than in Burke’s
formulation.
This view is clearly compatible with democratic party government, pro-
vided that its exponents accept the legitimacy of opposition, whether on
philosophical or prudential grounds (see Katz 1997: 35). In more doctrinaire
forms, however, it is also the populist position (and the technocratic, fascist,
communist, and theocratic positions as well). Simply, if there is an extrin-
sically identified “common good,” be it based on a hegemonic “political
logic” (Laclau 2005: 117), on impartial expertise deployed to “organize and
control . . . resources for the good of all” (Rosanvallon 2011: 48), on the
inherent nature of “the people” (das Volk) as determined by its leader (der
Führer), on historical materialism as interpreted by the vanguard of the
proletariat, or on the will of God as manifested through priests or ayatollahs
or rabbis, then both the mediation of conflict through party and indeed
opposition itself become fundamentally illegitimate. Hence in these forms,
the ideal of a unitary common good represents a direct challenge to the
model of democratic party government.
Uneasily coexisting with this communitarian view is the pluralist or liberal or
“economic” view according to which there are many competing and independ-
ently legitimate private interests instead of a single national interest. From this
perspective, the only possible definition of a national interest is the sum of the
private interests of the individuals comprising the nation—and given the appar-
ent impossibility of making interpersonal comparisons of utility, even that must
remain essentially contested. Because there is no existential national interest, no
party (or other political actor) can legitimately claim to represent it. Instead, a
party can only aim and claim to represent one or more of the particular interests
in society, in contrast to and in competition with other interests. As the
explicitly class or religiously based parties of the late nineteenth and early
twentieth centuries softened their appeals to become catch-all parties
(Kirchheimer 1966), their role as broker/mediator among interests, and
between interests and the state, continued to imply that there are interests (in
the plural) that need to be accommodated rather than needing to be dissolved
into a singular and homogeneous whole. This is clearly compatible with the
model of party government—and not with the model of populism—because
from this perspective, the focus is not on the selection of the “right” government
that can represent the singular people, but rather on the selection of represen-
tatives who can collectively represent and accommodate a diversity of interests.

The problem arises when citizens expect parties to act in accordance with
both views at the same time, and criticize parties for pursuing sectional
interests (“special” interests when they refer to someone other than the critic)
while at the same time working within political institutions and theoretical
understandings predicated on the idea that those sectional interests are the
only kind of interests that there are. On the one hand, parties are expected to
represent the personal and sectional interests of their supporters on pain of
losing that support and leaving those supporters feeling unrepresented, but on
the other hand they are expected to rise above sectarianism and partisanship
more generally to govern in the national interest. The public demands dem-
ocracy, but has a visceral dislike of politics. In a telling use of language, for a
politician to be described as acting “politically” has become a criticism.
This, then, intersects with incompatible conceptions of the proper nature of
political parties in democracies: on the one hand, the ideal of parties as
principled and internally democratic organizations of citizens19 and, on the
other hand, the ideal of responsible teams of leaders ultimately converging on
a single middle-of-the-road national will or national interest.20
From the first perspective, one returns to the ideal of the membership-based
party that was characteristic of the mass party of integration—and the idea of
class- (or other group) based representation. The party in public office is the
agent of the party on the ground; the members of the party are not merely its
supporters, they are the defining essence of the party itself, and hence they are
the party’s authoritative policy makers. Of course, as many party scholars,
following from the analysis of Robert Michels (1962 [1911]), have observed,
the principle of internal party democracy generally is more honored in the
breach than in the observance, but this is, in fact, the point. While the party in
central office may have become the master rather than the servant of the party
on the ground, and while the party in public office may come to dominate
both (Chapter 3; see also, Katz and Mair 2002), this continues to be regarded
as a pathology, and failure to provide effective avenues for membership
control over party policy, strategy, and personnel is regarded as grounds for
criticism (Cross and Katz 2013; see also Gauja 2015).
19
For examples, see Chapter 2, note 13. Here we ignore the likely incompatibility of internal
democracy and the appearance of principled consistency in a multi-dimensional policy space that
is implicit in Arrow’s impossibility theorem. See also Downs (1957: 25).
20
This also clearly raises Mair’s (2009) concerns with the distinction between representation
and responsibility. An earlier version of essentially the same dichotomy, albeit with a quite
different evaluation, was expressed by L. S. Amery (1947: 30–1) reflecting on the meaning of
“responsibility” to British Tories: “The word ‘responsibility’ has, however, two senses. It connotes
not only accountability to an outside or final authority. It also connotes a state of mind, which
weighs the consequences of action and then acts, irrespective, it may be, of the concurrence or
approval of others.”

The alternative perspective can perhaps most clearly be seen in thinking
about the place of parties as institutions in a democracy that is derived from
the Downsian model of convergence to the median voter (Downs 1957;
Mueller 1989). The appeal of the median voter theorem in normative terms,
and of moderation in practical terms, is obvious—at least to moderates. It is
easy to forget, however, that the expectation of convergence depends on
parties not being internally democratic: either to allow diversity of opinion
among the party’s leaders, or to allow “multitudes of other citizens also [to]
take part in [party] decision-making,” implies that “the actions taken by the
party as a whole are likely to form a hodgepodge of compromises—the result
of an internal power struggle” (Downs 1957: 25) rather than a coherent
position (that can be modeled as a single point on a single dimension) chosen
solely in the expectation of maximizing the probability of electoral victory.
It is also easy to forget that the other crucial assumption of the Downsian
model is that politicians should be utterly self-interested, and personally
indifferent with respect to policy. It is precisely because they are without
fixed political principles that parties are free to respond to the demands of
the public. The situation appears a bit more complicated in the case of
parliamentary coalitions, but in essence it is the same. Although parties may
not be expected to converge in the electorate,21 they have to accept the price
of appearing to the public to have abandoned their campaign pledges if they
are to be able to form a governing coalition, leaving them as apparently
unreliable agents of their electors.
Again there are radically inconsistent standards of evaluation, but an
expectation that parties should satisfy both. Even within the responsible
two-parties model of democracy, one finds what Austin Ranney called the
“little civil war” about internal party democracy (1962: 156). In a multi-party
system, the German party law, for example, requires that parties be demo-
cratic in their organization, not just in their support for democracy at the
system level. But at the same time the German basic law says that members of
the Bundestag “shall be representatives of the whole people, not bound by
orders and instructions, and shall be subject only to their conscience”22—even
as the electoral system means that voters are choosing parties rather than
individual MPs.23 On one side, commentators admire “profiles in courage” (as
21
In fact, although the Downsian argument for convergence depends on there being exactly
two parties, the expectation that parties should converge, even in multi-party systems, remains
quite common.
22
See Chapter 5, note 1.
23
And while both of these versions of “responsible party government” assume coherence
within parties—so that they can be held accountable—citizens continue to want representatives
who will support local (or other particular) interests in defiance of their parties. See, for example,
Pearse (2005).

John Kennedy (1957) described a few American senators who did what they
thought was right, even though it was unpopular)—but then promptly com-
plain about their undemocratic arrogance. On the other side, they complain
about “Apologetic statesmen of a compromising kind,” as W. S. Gilbert’s
Lord High Executioner in The Mikado described them in his list of people
who could conveniently be executed because “they’ll none of ‘em be missed—
they’ll none of ‘em be missed,” but then praise the statesmanship of the
coalition builder. Put in the terms of political science rather than political
commentary, this reflects Strm’s (1990) conflict between policy and office as
goals for political parties.24
THE RISE OF POPULISM
Although the tensions between preference and expertise in policy making,

between communitarian and liberal views of society, and between member-
driven and elite-driven parties are not identical, they are strongly related. To a
great extent, the governing as technique view goes with the single-national-
interest view, which goes with the flexible and autonomous party leadership
view. Conversely, the policy-making-as-preference-implementation view goes
with the internally democratic-party view which goes with the multiple-
societal-interests view. As a result, the problems stemming from these conflicts
tend to reinforce one another, and this mutual reinforcement is especially
relevant for the problem of populism.
Entering the second decade of the twenty-first century, a number of things
had occurred that have made the mainstream parties particularly vulnerable
to challenges catalyzed by the crises of immigration, the burden of public
debt, the global financial crisis of 2008 in general and the eurozone crisis in
particular (especially in Greece, but to a lesser extent in other European
countries), and the resulting rise in Euroskepticism. Detachment from parties,
and more generally atomization of society, have left many people without
strong feelings of belonging, something that nationalists can provide given a
suitable “other”—which has been supplied in Europe by the influx of foreign-
ers from other EU countries and especially by the influx of refugees from
24
Strm is unclear as to whether “policy” means having an impact on the ultimate policy
adopted by the state (for which achieving office would be a means rather than an alternative), or
maintaining steadfast adherence to the advocacy of a policy position. Clearly in this paragraph we
are assuming the latter position.

Africa and the Middle East, and in the United States by the large numbers of
undocumented Latino immigrants, and by the assertiveness of African Ameri-
cans unwilling to accept permanent second-class status (made all the more
threatening to “white” America by the election of Barack Obama as president
of the United States). The increasingly tight restrictions on party policy—
whether imposed by reality or by choice, and whether imposed on the parties
or by the parties—have fostered the appearance of parties as agents of an
interparty, and indeed international, elite combination of governments, bank-
ers, and globalized business (embodied in “the state,” the EU, the World
Bank, and IMF, WTO, etc.) rather than agents of society working to control
those forces in the interests of ordinary citizens. This then resonates with the
populist belief that the “powers that be” are in a conspiracy to protect
themselves at the expense of the people. Even though the crises were not a
direct consequence of cartelization, they have been exacerbated by the reac-
tions, or non-reactions, of cartel parties, leading citizens increasingly to be
receptive to the complaint that the mainstream parties have become part of
the problem rather than being part of the solution, a receptiveness that is then
furthered by the perception that the parties are “feeding at the public trough”
without actually adding much value in return. (See, for example, Foxe 2011;
Stott 1995—both entitled “snouts in the trough,” albeit about different coun-
tries and with correspondingly different subtitles.)
All of this is then reinforced by the tendency of the news media to focus on
happenings and debates in the capital, and to rely for analysis on the same
relatively closed circle of experts that are advising the parties. One result is to
further the self-referential nature of elite political discourse, effectively
generalizing to the entire mainstream party system the complaint levied
against the French National Assembly under the Fourth Republic, that it
had “become a ‘house without windows’ playing an endogenous parliamen-
tary game indifferent to the environment outside” (Patterson and Copeland
1994: 6). One manifestation of this may be seen in the media’s tendency to
report, and then the political elite’s tendency to interpret, the failure of
populist candidates such as Geert Wilders or Marine Le Pen to win their
elections as reassuring victories for the “responsible” parties, rather than
reporting and interpreting the populists’ significant electoral gains as cause
for concern. A second result is to leave large numbers of citizens with the
(often quite accurate) sense that they and their needs are not being paid
much attention.
Against this background, we can understand both the widespread turn
away from the traditionally “responsible” parties in favor of “populist”
movements, and why this might be characterized as a conflict between
democracy and technocracy, even as the “responsible” democratic parties
turn to technocracy as a refuge to maintain their own positions. Ironically, in
this conflict it is the populist pole that sometimes comes to be identified with

democracy, notwithstanding the widespread idea that populism is, itself, an
anti-democratic force.25
All of this has the potential to create a crisis of legitimacy, but legitimacy of
what? It does not appear that the legitimacy of “democracy”—or perhaps
more specifically, the legitimacy of government by elected representatives—as
an ideal is currently in question. The same thing cannot be said with equal
assurance with regard to the legitimacy of the current party systems, or even
with regard to the legitimacy of party government as a way of institutional-
izing democracy. In particular, if the people come to see the alternatives
offered by the ostensibly competing parties simply as choices among different
members of the same gang because after throwing the scoundrels out, the new
scoundrels do the same thing as the old scoundrels, then reduction in the
diffuse support for the core of the party system, and not just for the particular
parties that happened to be in office, is likely to result.26 The spectacular rise
of populist parties in countries like Greece, Italy, and Spain can certainly be
understood as evidence of this in response to reforms seen to be imposed by
technocrats and Eurocrats, and then acceded to by “responsible” politicians,
but similar, if weaker, trends are visible elsewhere.27 In Ireland (an early
“victim” of troika-required austerity), for example, although the parties in
government in 2007 (Fianna Fáil, Greens, Progressive Democrats28) saw a
25
The root of this apparent contradiction lies in the ambiguous and multiple definitions of
democracy. On the one hand, democracy suggests that ultimate power must rest with the demos;
democracy implies that the “will of the people” ought to determine public policy—and here it is
perhaps relevant to observe that the word demos, in line with the populist view of a homogeneous
“people,” is a singular noun. Technocracy, in suggesting that experts—and especially experts who
are outside of the demos—rather than giving advice that the people can take or reject as the people
decide, should have the ability to impose decisions against the will of the people, is clearly an
alternative to, or at least a limitation of, democracy. In any event, as Bickerton and Invernizzi
Accetti (2017) as well as Caramani (2015) point out, in denying the essential centrality of party
and representative institutions to democratic politics, both populism and technocracy are
antithetical to democratic party government.
On the other hand, at least in the modern world, democracy has become identified with the
protection of rights, not just as a necessary pre-condition for popular rule but as coequal, or even
superior, to popular rule in the identification of democracy. In this respect, the tendency of
populists to scapegoat outsiders, whether defined by nationality or ethnicity, qualifies as a
significant challenge to democracy.
26
Two examples of this from the United States would be the Cato Institute commentary “Why
Vote for Tweedle-Dum or Tweedle-Dumber?” (Bandow 1998) or George Wallace’s claim in 1968
that “There ain’t a dime’s worth of difference” between Republicans and Democrats.
27
Although the Italian upstarts did not win in 2013 to the same degree as the Greek upstarts
won in 2015, the fact that the Italian election was held at a time when all of the main parties were
supporting the caretaker government of Mario Monti probably contributed to the success of the
upstarts by lowering the salience of distinctions within the mainstream, and thus allowing
discontent with the mainstream to become more salient (Segatti 2014: 134).
28
Disbanded in 2009.

hemorrhage of votes to the erstwhile opposition of Fine Gael and Labour,
the five “mainstream” parties taken together saw a loss of over 11 percent of
the vote.
Evidently, many Irish voters did not see a shift from one of the major
parties to another (or even to one of their smaller partners) adequately to
reflect their discontent with the performance of the government in the wake of
the banking crisis. In the Irish case, with the single transferrable vote electoral
system, many of these voters could transfer their support to independents. In
many other countries, however, this option is not seriously available. Thus, to
the extent that all of the “responsible” parties are seen to be implicated in
effectively the same set of economic (or European) policies, the only alterna-
tive for those who want to express their rejection of those policies is to vote for
an “irresponsible”—in this case, populist—party, notwithstanding that the
voters may not accept many of the less savory elements of the party’s pro-
gram. Moreover, even voters who approve of the policies of the mainstream
parties are likely to be less enthusiastic in their support, and less likely to turn
out to vote, if they do not perceive those policies to be subject to any
substantial challenge.
This collapse of support for the erstwhile mainstream has not been limited,
however, to the countries that were most directly affected by the recent
economic crises. As Table 7.5 shows for Austria, Germany, the Netherlands,
and the UK, the total share of the popular vote won by the parties that were
the core of the mainstream in their countries in the 1960s has dropped
dramatically since then. Moreover, if the numbers were calculated as a
share of the electorate rather than of the votes cast, thus taking into account
the significant drop in turnout in these countries, the decline in support for the
old mainstream would be even more stark. While some of this drop has been
compensated by the cooptation of new parties into the mainstream, it only
reduces, but does not reverse, the collapse. For example, even adding the votes
of both the FDP and the Greens to the SPD+CDU/CSU total leaves the
augmented German mainstream in 2017 with a smaller percentage of the vote
than the two parties alone had at any election before 2002.
T A B L E 7 . 5 Average vote share of the mainstream parties by decade: Austria,

Germany, the Netherlands, and the UK
1960s 1980s 1990s 2000s 2010s
Germany (SPD+CDU/CSU) 86 85 77 68 60
Austria (OVP+SPO) 91 84 60 55 51
Netherlands (PvdA+VVD+CDA) 83 82 69 65 51
UK (Labour+Conservative) 90 72 75 70 72

Although the delegation of authority to “experts,” most commonly bankers
and economists, may have been originally justified as insulating important
areas of policy from partisan politics (which might be expressed more pejora-
tively as allowing governing parties to avoid responsibility for those policies
and their results), the very denial of responsibility undercuts the parties’
claims to respect and privilege.29 If the experts produce acceptable results (or
at least if the results are acceptable, whether or not the experts have a legitimate
claim to responsibility for them), this is unlikely to cause problems—with
objections coming, if at all, only from those who could be identified as “irre-
sponsible cranks.” But as with the so-called permissive consensus regarding
Europeanization, when the results became less good than the people had come
to expect (at least in part because of the overpromising of the established parties),
or when economic pain was imposed in the name of meeting the targets of the
Stability and Growth Pact (which for some countries appears to have produced
neither), and that was coupled with an influx of immigrants, not only the process
of Europeanization but the degree to which the governing parties could be relied
upon to defend the interests of their own citizens came under more widespread
challenge, setting up the conflict between populists who claim to speak for the
people and the mainstream parties, which appear to be in the thrall of Eurocrats,
technocrats, and other experts—and whose leaders may themselves appear to
party members to be manifesting the arrogance of technocrats.
LOOKING TO THE FUTURE
As Herbert Dőring observed in a related context, there are two ways to deal
with disappointed expectations. One is to improve performance, but the other
is to lower expectations (Dőring 1987: 149). Obviously, one way for the gap
between expectations and performance to be narrowed would be for perceived
substantive performance to improve. While this might have little bearing on
the gap between expectations and the democratic performance of parties, it
would be likely to lower the salience of the whole question—at least until the
next economic or social crisis. While such improved performance is certainly
possible, even the experts claiming that the current pain of extreme austerity is
necessary to secure growth in the future seem to believe that in the case of
countries like Greece this growth will only come after a number of years.
Similarly, the current influx of immigrants seems unlikely to be abated at any
29
Although it is easy to see the short-run incentive to claim that problems are “not my fault/
responsibility” rather than to admit having made a mistake, in the longer term, if outcomes are
not the responsibility of the parties, what are the parties good for?

time soon. The question, then, is whether improved performance is possible,
and even if it is possible, whether it will come before significant long-term
damage is done to the legitimacy of party democracy. But there is another
way in which democratic performance could be improved, even if economic
(or other policy) performance lags. That would be for the apparent consensus/
conspiracy among the mainstream parties that austerity is necessary, that the
tide of immigration cannot or should not be stemmed, that Europeanization
must proceed, etc. to be broken. Many voters still might not get the policies
they want, but at least they would be able to express their frustration without
having to turn to the populist extremes.
There are some signs that this may be possible in the selection of Jeremy
Corbyn as leader of the British Labour party and the resulting move to the left
of the party’s program. Although Corbyn’s selection in September 2015,
followed by an overwhelming vote of no confidence by the Parliamentary
Labour Party in June and then his confirmation as leader in September 2016
in an all-member (plus supporters) vote, appeared to doom the party (Labour
polling support that had hovered around 35 percent into March 2016 fell to
around 27 percent in June with no sign of recovery), by the general election
of 2017 there had been a remarkable turnaround, with Labour winning
40 percent of the vote and gaining thirty seats in the House of Commons (as
against 42.4 percent of the vote and a loss of thirteen seats for the “victorious”
Conservatives). Whether these gains by Labour are the result of the party’s
more differentiated policy image, and therefore suggestive of the possibility of
containing the rise of anti-party-system party support by offering more mean-
ingful choices within the mainstream, or alternatively the result of infighting
and ineptitude among the Conservatives, remains to be seen. It also remains
to be seen whether the dramatic increase of Labour Party membership (from
190,000 full members in May 2015 to 515,000 in July 2016 (Stewart
forthcoming)) will prove indicative of a revitalization of the party or merely
of a temporary influx of “tourists” and “gate crashers” (Stewart and Archer
2000). Moreover, even if the increase in membership proves durable, it is less
likely to reflect an increase in the power of the party on the ground, and more
likely to reflect a different mode of competition within the party in public
office for control over the party on the ground: in effect, Jeremy Corbyn
“recruited a new party to outvote members of the old one” (Rentoul 2017).
There are also some signs that the mainstream consensus may be changing
to be less welcoming of immigration (for example, Angela Merkel’s October
2017 acceptance of limits on the acceptance of refugees—Eddy 2017). Even
more, there is the flip of both Labour and Conservative parties in Britain from
opposition to Brexit before the 2016 referendum to support of Brexit in the
2017 general election. While both of these can be interpreted as showing the
power of public pressure to influence the policies even of cartel parties, two
points need to be emphasized. The first is that, as in the Downsian competitive

model, cartel parties use policy as a means to the end of securing their own
advantage—in this case by limiting competition—rather than as an end in
itself. To change policy as a result of losses (either share of votes cast or share
of the electorate) for the cartel as a whole (see again Table 7.5) is entirely
consistent with cartelized behavior. The second is that when the whole main-
stream shifts position together, it may be responding to aggregate electoral
opinion, but it is not increasing the range of options offered.
The other possibility is for expectations to be brought more into line with
performance. And again, this has both an economic and a political aspect, but
in this case as complements rather than as alternatives. On the one hand,
people might become more accepting of the constraints under which govern-
ments claim to be working—in effect (in the words of John Stuart Mill)
“accepting [the] results, when disagreeable, as visitations of Nature” (Mill
1982 [1861]: 52)—with “Nature” being a euphemism for forces that are
accepted to be beyond human control. Certainly levels of unemployment
(much lower than currently experienced in much of Europe) that were once
thought to be guarantees of revolution have come to be accepted as normal,
and perhaps even levels approaching those currently experienced will eventu-
ally be accepted as the result of conditions beyond the control of any demo-
cratic government. But if, on the other hand, the economic well-being of the
people is accepted as being beyond the control of democratic government,
then undemocratic alternatives may see a real (re)surgence. Alexis Tsipras’ re-
election in September 2015 after accepting, however reluctantly, what
amounted to EU suzerainty, may prove the crucial test case in this respect.
Syriza lost votes (0.8 percent) and seats (four) and turnout dropped by
7 percent (to the lowest figure since the restoration of democracy in 1974) in
comparison to the January election, but it benefitted from the facts that the
new party formed by internal opposition to Tsipras did not have enough time
to mount an effective counter campaign while the two main opposition parties
were undergoing leadership changes (Mylonas 2016: 116). By the end of
November 2016, however, polling support for Syriza had dropped to below
20 percent (recovering to the low 20s in mid-2017), while support for New
Democracy was in the high 30s. Whether or not New Democracy’s proposals
are realistic, it appears that Tsipras’ argument that there was no alternative
to accepting the demands of the EU and IMF had limited staying power.
On the other hand, support for the ultranationalist Golden Dawn only
increased modestly.
An alternative version of ratcheting down expectations might be suggested
by Yishai’s (2001) idea of “post-cartel parties” in Israel. In part driven by the
limitations of cartelization, particularly with a hyper-proportional electoral
system, and in part by a shift in power “from the legislature to the adminis-
tration, and particularly to the Supreme Court [she suggests that] political
parties resumed their linkage with civil society.” Whether this has proven a

successful strategy is, however, open to question. Five years after Yishai’s
article was published, Israeli election turnout reached an all-time low
(63.5 percent), and while it has since risen to 72.4 percent, this is still lower
than in any election before 2000. The vote share of the two biggest parties
(Labor and Likud), which had been steadily declining (from 59.6 percent in
1992 to 34.5 percent in 1999—although that was influenced by the possibility
of voting for a major party candidate in the direct election of the prime
minister while voting for another party for parliament) “recovered” to 43.9
percent in 2003 and 46.1 percent in 2006 (including Kadima as well as Likud),
but fell to only 42.1 percent in 2015. Moreover, with ten lists (becoming eleven
party groups in the Knesset representing seventeen parties), and a pattern of
coalition formation based as much on the buying of support in exchange for
policies favoring particularistic interests as it is on ideological affinity, Israel
does not appear so much to represent a move towards more mass-party-like
democratic politics, as a move towards pre-democratic monarchic politics
(albeit now with a democratically elected “king”), in which parties voice
grievances and try to extract benefits in exchange for diffuse consent.
Rather than a dichotomy of pluralist party democracy versus populism,
one might posit a trichotomy of populism, technocracy, and party democracy.
What we have been seeing over the recent decades is a growing alliance
between the mainstream parties and technocracy (under such labels as “the
regulatory state”)—in effect, acceptance by all of the mainstream parties of
the Burkean idea that it is better to be governed well by a qualified elite, now
defined by technical training rather than by birth, than it is to have self-
government. So long as things went well—so that popular tastes were being
satisfied by experts’ techniques—the parties could maintain their privileged
social position while the technocrats managed societal governance and the
populists had little basis for support beyond blatant racism or xenophobia.
When things ceased to go well, however, the parties found themselves as the
most visible targets for dissatisfaction. From the populist side, they had
failed to attend to the interests of the people, instead protecting “the
establishment.” From the technocratic side, they were too quick to bend
to political pressure. The result has seen increasingly prominent calls to
bypass the politicians, either by granting ever more independent authority to
the technocrats or by replacing them with “amateurs” at the polls.30 Both
30
Although it might be argued that the United States is a special case, this point is illustrated
by a September 2015 CNN poll showing an absolute majority of US Republicans to support one
of the two leading contenders (Trump and Carson) for the Republican presidential nomination—
neither of whom had ever held elective office at any level. Later in the month, although support for
Trump appeared to decline somewhat, the percentage supporting a candidate with no experience
in elective office increased, as Trump and Carson were joined at the top of the polls by the equally
inexperienced Carly Fiorina. Ultimately, of course, Trump won both the nomination and the

represent serious challenges to party government as the currently dominant
embodiment of democracy and, if as Schattschneider (1942: 1) said over
seventy years ago, “modern democracy is unthinkable save in terms of the
parties,” then to democracy itself.
A FINAL NOTE
Although it is not clear that democracy really is in peril, the growth of

populism (in particular extreme right populism) in Europe and the election
of Donald Trump as president of the United States have certainly led some to
that conclusion. Even though they rarely, if ever, use the term “cartel,” it is
evident that the growing separation of citizens and parties that is one of the
core elements of the cartel party model is also one of the core claims of the
populist, anti-party-system-party politicians. To the extent that they still
depend on private contributions, mainstream politicians are responsive to
their big financial backers. They are responsive to technocrats, both within
their own countries and in international institutions like the EU or the
WTO. They are responsive to one another. But they are not responsive to
ordinary citizens. The natural question (pace Lenin) is “what is to be done?” If
the problem stems from the cartelization of the mainstream parties, perhaps
the answer is to reform those parties.
Former New York state governor (1919–20 and 1922–8), and 1928 Demo-
cratic Party candidate for the presidency of the United States, Alfred
E. Smith, is widely cited as saying “If there are any ills that democracy is
suffering from today, they can only be cured by more democracy,”31 although
some form of the claim clearly predates that statement. While never without
critics, either then (see, for example, Implement Record 1922)32 or now
(Brennan 2016), the idea that one way to address the current “crisis of
democracy” is to increase the internal democracy of political parties clearly
has been latched onto by many party scholars and party leaders. In earlier
work (Katz 2013, and reflected at various places in this volume), we have
already suggested why it may be too optimistic to believe that increased internal
party democracy (or “discounted” or “multi-speed” party memberships) will
White House, with his lack of political experience (put more positively as his lack of entanglement
with the establishment elite of either party) seen as an asset by many voters.
31
Quoted in “Smith would give more power to city,” New York Times, February 4, 1923,
https://timesmachine.nytimes.com/timesmachine/1923/02/04/105845183.html?pageNumber=59.
32
For example, “The popular slogan of today that the cure for the ills of democracy is more
democracy is no more true than to affirm that the cure for indigestion is more food.” Implement
Record: Tractors and Farm Equipment 19 (November 1922), p. 34.

prove effective long-term fixes to the near universal decline of party member-
ship or popular alienation from parties more generally.
Aside from the idea that increased popular participation in political parties
would be good for its own sake, increased internal democracy generally is
supported on the assumption that it will force party leaders to be more
responsive. The obvious question that must be asked here is “responsive to
whom?”, and in particular whether partisan officials should be responsive
primarily to the general public or to their own party’s members or activists—
or their financial supporters. And if responsiveness is to the party’s members
or activists, two other questions arise: On the one hand, are party activists or
members motivated or equipped in ways that should lead us to want politi-
cians to be especially responsive to them? And on the other hand, what are the
likely consequences for the behavior of politicians in forcing them to be
particularly responsive to members or activists?
In some ways, the United States represents a very good venue to address
these questions, although in other ways the relevance of the American case to
other systems—and particularly to parliamentary systems—can at least be
questioned. Certainly, with regard to selection of candidates, American parties
represent the extreme on empowerment of the base, although it is important to
remember that this empowerment is largely imposed by law rather than party
rules, and was originally intended to weaken the party organizations, not to
strengthen them. Moreover, this empowerment has to be understood in the
context of an extremely amorphous and porous understanding of what the
party base is.33 As a result, nomination decisions are made by primary
(or caucus) voters who are often without strong organizational ties to “their”
party (the equivalent of Stewart and Archer’s (2000) gatecrashers and tourists).
But as an increasing number of parties in other countries open participation in
their leadership- or candidate-selection processes to “registered supporters”
and the like, the uniqueness of the American system is clearly being reduced.
Although voter turnout in the United States is low by international stand-
ards, primary election turnout is even lower, only (barely) exceeding 30 percent
in one presidential contest since 1980, and generally closer to (and often below)
20 percent; primary turnout in non-presidential years is even lower.34 And,
33
Here we use the term “base” to refer to the nearest American equivalent to “membership” in
parties in most other countries. In this sense, the party base is generally identified with those who
have chosen a party designation in the process of registering to vote, but not every state has
partisan registration, and partisan registration may indicate a desire to be able to vote in the
party’s primary elections rather than real support for the party. This should not be confused with
the more typical American usage of the term to refer to hard-core contributors and supporters.
34
William A. Galston and Elaine C Kamarck calculated turnout in the highly contested 2010
Congressional primaries to have averaged 7.5 percent. https://www.brookings.edu/wp-content/
uploads/2016/06/KamarckIncreasing-Turnout-in-Congressional-Primaries72614.pdf, accessed
November 10, 2017.

of course, primary voters are not a random sample of the electorate, but
instead significantly overrepresent the most partisan of the party’s sup-
porters (with “partisan” referring more to ideological dogmatism or hostility
to the other party than to a concern for the long-term interests of “their
own” party). It is worth observing here that although May’s Law is generally
expressed in terms of the relative extremism of activists as opposed to voters,
in some passages the word May uses is “intransigence” rather than extrem-
ism (May 1973: 141).
This system forces candidates to be independent operators in raising their
own campaign resources to defend against primary challengers—or ideally to
scare them off before they mount a challenge in the first place—and therefore
makes candidates especially dependent on those who can provide those
resources.35 Essentially by definition, it increases the personalization of pol-
itics. It also means that before they can try to be responsive to the general
electorate, would-be office holders must first respond (or pander) to the most
extreme and intransigent of their party’s supporters, the small minority who
actually vote in primary elections. Moreover, the potentially pernicious
impact of this minority is only increased by the anonymity with which their
participation takes place, inviting some primary voters to privilege their
private preferences (“Who would I like to see elected?”) over their assessments
of the party’s interests (“Which of the aspirants to the nomination is most
likely to be able to win the general election?”).36
American experience suggests three potential negative consequences of
this. The first is a reduction of the incentives for party cohesion. Politicians
who have raised their own resources, built their own support coalitions, and
won the right to their party’s label in a primary election have far less incentive
to defer to their party leaders than they would if those leaders had the ability
to withhold these necessities. Moreover, while personalism is naturally stron-
gest in primary elections, where there are no party labels to distinguish the
candidates, the focus on the candidates as individuals carries over into the
general election, reducing the relevance of party, and encouraging candidates
to maintain profiles separate from, and even in opposition to, their parties.
Although party unity in Congress has increased dramatically since the time
35
In the American context, this generally means the super-rich, but it also includes groups that
are able to mobilize highly committed supporters—of the group, not necessarily of the party.
36
This is not to say that some Democrats in 2016 voted for Hillary Clinton rather than Bernie
Sanders in the (perhaps mistaken) belief that she was “more electable.” In the 2017 French
presidential election, both Socialist and Republican primaries were won by candidates who
were popular within their own parties, but not with the electorate as a whole (neither advancing
to the second round). In 2012, on the other hand, Socialist primary voters chose François
Hollande over Martine Aubry, apparently in the belief that he was more electable, although
less “socialist.”

when a Democratic senator could vote with a majority of Republicans more
often than the average Republican, be keynote speaker at the Republican
national conventions, endorse Republican candidates for office, and yet
remain a member of the Senate Democratic Conference, it is still the case
that even a party controlling both houses of Congress and the presidency may
be unable to enact its policies due to internal defection on key votes. As a
result, while individual representatives may be held accountable for their own
votes by their individual electorates, it is far more difficult for their electorates
to hold them accountable for outcomes or for the electorate as a whole to hold
the parties accountable for anything.
The second consequence is to reinforce the incentives for extremism,
intransigence, and attempts, even by long-time office holders, to assume the
mantel of “outsider” and to run against the “establishment.” Democratic
politics requires compromise, both within and often between parties, but the
fact that the threat of being defeated in a primary is much greater for most
incumbents than the threat of losing in the general election,37 means that
those who wish to retain their seats (as Republican senate leader Mitch
McConnell said, “winners make policy, losers go home”), often forces incum-
bents to be especially attentive to those elements of their party’s electorate
who are most extreme in their views, and least willing to forgive deviation
from their own preferences. And this in turn encourages a strategy of deflect-
ing blame/responsibility by, for example, “running for Congress by running
against Congress” (Fenno 1978: 168; for a suggestion that this has become less
universal, see Lipinski et al. 2003). Ironically, even if this strategy is successful
on a personal level, it also undermines public confidence in the institutions of
government and thus reinforces the populist claim that those institutions are
fundamentally corrupt.
Once the primary is over, the candidate who was successful in appealing to
his/her party’s primary electorate then has to appeal to the general electorate,
which includes not only the less committed supporters of his/her own party,
but independents and partisans of the other party or parties. Even the median
registrant of the candidate’s party (and the primary electorate is likely to be
even more extreme than the median of the party overall) is certainly going to
be distant from the overall electoral median. This means that the electorally
dominant policy positions for the general election are likely to be quite
different from those taken in the primary. This leaves the candidate with
two strategies. On the one hand, the candidate can try to moderate (“walk
back”) the primary positions to make more centrist appeals—at the cost of
37
Writing about state legislatures, for example, Seth Masket (2009: 11) argues that
“legislators . . . fear the wrath of the activists and other political actors who control the primaries
more than they fear the judgment of voters in the general election.”

appearing inconsistent, insincere, or simply opportunistic. On the other hand,
the candidate can try to avoid firm or clear positions in either election, for
example appealing to and exacerbating “tribal” loyalties and antagonisms, at
the cost of appearing to be without principles. In either case, the result is again
effectively to confirm the anti-party-system populist view.
Although the United States is an extreme case, some elements of this
analysis find resonance in European parliamentary systems as well. No par-
liamentary democracy employs primary elections that are exactly the equiva-
lent of those in the United States, but there is an increasing trend toward
opening candidate- and leadership-selection decisions to all member votes
(and also to redefine membership in ways that come close to the “open
primary” electorates of the United States). Moreover, even without primary
elections, there is a growing tendency in list proportional representation
systems to increase the role of the party’s voters in determining which candi-
dates from the list actually are elected (Renwick and Pilet 2016). When
intraparty preference voting is possible, it may play a larger role than the
interparty result in determining which incumbent members of parliament are
defeated (Katz 1986) and which candidates, whether incumbents or new-
comers, are elected (Renwick and Pilet 2016: 243). Intraparty preference
voting also appears to have a detrimental impact on party cohesion, with
for example Finland, which has one of the most personalized PR systems in
Europe also showing the lowest levels of parliamentary party unity (Depauw
and Martin 2009).
One reasonable interpretation of calls for more internal party democracy is
that it reflects nostalgia for an imagined golden age of the mass party. In this
respect, it is worth recalling the adage of Will Rogers that “things ain’t what
they used to be and probably never was.” And also to recall that one major
thrust of research in the decades following World War II was to call into
question both the capacity of ordinary voters (e.g., Berelson et al. 1954) and
their commitment to liberal values (e.g., McClosky 1964). Translated to the
first decades of the twenty-first century, with the rise of populist movements
and widespread belief that one’s political opponents are filling the public
space with, and themselves believing and acting on, “fake news,” it may
well be that the problem is not too little internal party democracy but too
much, and that far from advocating more internal party democracy, advo-
cates of democratic party government should remember another adage attrib-
uted to Will Rogers: “If you’re in a hole, stop digging.”
If not more democracy, then what? The answer would appear to be a
combination of three elements:
(1) More responsible, in the sense of more moderate and deliberative,

behavior on the part of political elites coupled with willingness to
accept responsibility for decisions and correspondingly less emphasis

on elite responsiveness to, and exploitation or incitement of, ill-
informed raw public opinion.
(2) More attention to the building and maintenance of an inclusive political
community, and more community-oriented (socio-tropic) as opposed
to individualistic thinking on the part of citizens, coupled with a media
environment that better distinguishes fact from innuendo and encour-
ages attention to dialogue rather than isolation in self-verifying echo
chambers.
(3) More realistic expectations regarding the capacities of government, the
compromises necessary for coexistence in diverse communities, and the
inevitable weaknesses of human beings.
Unfortunately, the current short-run incentives for virtually every actor in the
political space run directly counter to these objectives. That this is the case
with regard to political elites has been an implicit, and often an explicit, theme
of much of this book. It is probably unrealistic to expect voters to distinguish
fact from fiction, or rationality from appeals to prejudice, without guidance
from a responsible information industry, but the annexation of journalism
by the entertainment industry and the pressure to make information a profit
center rather than a public service argue in favor of sensationalism, while the
increasing use of the internet for political information both eliminates many
of the peer-pressure social controls that encouraged journalistic responsibility
in the past, and facilitates retreat into homogeneous and prejudice-reinforcing
bubbles.
All is not necessarily without hope, however. Much of what is being said
about cartel parties by their anti-party-system challengers reflects complaints
raised by nineteenth-century socialists against the bourgeois “cartel” of elite
parties,38 while the bourgeois fear of the collapse of liberalism before the
socialist onslaught reflects much of the anguish of the current mainstream
when confronted by the current rise of populism. Projecting from this, one
might look for a split in the populist movements akin to that which occurred
in the socialist movement, with social democrats moderating both their
demands and their rhetoric and ultimately becoming part of the mainstream.
And there is already some evidence of that in the split on the left between the
Realos and the Fundis in the German Green Party (as well as in the Belgian
38
Even in the absence of a serious socialist party, consider the rhetoric of American William
Jennings Bryan’s speech at the 1896 Democratic convention: “Having behind us the producing
masses of this nation and the world, supported by the commercial interests, the laboring interests,
and the toilers everywhere, we will answer their demand for a gold standard by saying to them:
‘You shall not press down upon the brow of labor this crown of thorns; you shall not crucify
mankind upon a cross of gold.’ ” And in this context, it is worth remembering that in 1896 Bryan
was not only the candidate of the Democratic Party, but of the Populist Party as well—although
that was a populism of the left.

Flemish Agalev Party) in the 1980s and 1990s, and more recently on the right
in splits within the Austrian FPÖ and the German AfD. If nothing else,
parties like the Greens, the FPÖ, and the Italian Communists have all
discovered that it is very difficult to maintain a party’s anti-party-system
support while being in government (or even while being perceived to be
excessively cooperating with the government from outside). In this scenario,
with the new anti-party-system parties playing the role of the old mass
parties39 and the current mainstream or cartel parties cast in the role of the
old elite parties, we might look for a new synthesis. While this would be
entirely consistent with the model of evolutionary party development that is
central to the cartel party thesis, it is far too early to say what that new
synthesis is likely to be.
39
There are obvious differences, particularly with regard to claims of internal party
democracy, but it is not unreasonable to see the “encapsulation” of populist supporters in
internet-based echo chambers, and their general sense of grievance and being ignored or
demonized by the rest of society, as resulting in a twenty-first-century equivalent of the mass
party’s strategy of encapsulation.
References
Agius, Christine (2007) “Sweden’s 2006 Parliamentary Election and After: Contesting
or Consolidating the Swedish Model?” Parliamentary Affairs 24 (4): 585–600.
Agrimarketing (2015) “Estimated U.S. seed market shares released.” http://www.
agrimarketing.com/s/97703, accessed May 1, 2017.
Aldrich, John (1995) Why Parties? The Origin and Transformation of Political Parties
in America. Chicago, IL: University of Chicago Press.
Aldrich, John (2011) Why Parties? A Second Look. Chicago, IL: University of Chicago
Press.
Allern, Elin Haugsgjerd (2012) “Appointments to Public Administration in Norway:
No Room for Political Parties?” in Petr Kopecký, Peter Mair, and Maria Spirova
(eds), Party Patronage and Party Government in European Democracies. Oxford:
Oxford University Press, pp. 272–93.
Allern, Elin Haugsgjerd and Tània Verge (2017) “Still Connecting with Society?
Political Parties’ Formal Links with Social Groups in the Twenty-First Century,”
in Susan Scarrow, Paul Webb, and Thomas Poguntke (eds), Organizing Political
Parties: Representation, Participation, and Power. Oxford: Oxford University Press,
pp. 106–35.
Allum, Percy (1974) “Italy: Ungovernable Republic.” Europa-Archiv 29: 717–28.
Amery, L. S. (1947) Thoughts on the Constitution. London: Oxford University Press.
Ames, Paul (2014) “High-profile corruption scandals are rocking European politics.”
https://www.pri.org/stories/2014-12-04/high-profile-corruption-scandals-are-rocking-
european-politics, accessed February 7, 2017.
Andeweg, Rudy B. and Galen A. Irwin (2002) Governance and Politics of the Netherlands.
Houndmills: Palgrave Macmillan.
Apokoronasnews (2015) “Former Greek finance minister given suspended sentence
over ‘Lagarde list.’ ” http://apokoronasnews.gr/former-greek-finance-minister-given-
suspended-sentence-over-lagarde-list, accessed December 10, 2015.
Avnon, Dan (1995) “Parties Laws in Democratic Systems of Government.” Journal of
Legislative Studies 1 (2): 283–300.
Bachrach, Peter (1967) The Theory of Democratic Elitism: A Critique. Boston, MA:
Little, Brown.
Bäck, Hanna, Marc Debus, Jochen Müller, and Henry Bäck (2013) “Regional
Government Formation in Varying Multi-Level Contexts: A Comparison of Eight
European Countries.” Regional Studies 47: 368–87.
Bairoch, Paul and Gary Goertz (1986) “Factors of Urbanisation in the Nineteenth
Century Developed Countries: A Descriptive and Econometric Analysis.” Urban
Studies 23 (4): 285–305.
Ball, Stuart (1995) The Conservative Party and British Politics, 1902–1951. London:
Longman.
190 References
Bandow, Doug (1998) “Why vote for Tweedle-Dum or Tweedle-Dumber?” https://
www.cato.org/publications/commentary/why-vote-tweedledum-or-tweedledumber,
accessed October 16, 2017.
Barber, Benjamin (1984) Strong Democracy: Participatory Politics for a New Age.
Berkeley, CA: University of California Press.
Bardi, Luciano, Enrico Calossi, and Eugenio Pizzimenti (2017) “Which Face Comes
First? The Ascendancy of the Party in Public Office,” in Susan Scarrow, Paul Webb,
and Thomas Poguntke (eds), Organizing Political Parties: Representation, Partici-
pation, and Power. Oxford: Oxford University Press, pp. 62–83.
Bartels, Larry (1992) “The Impact of Electioneering in the United States,” in David
Butler and Austin Ranney (eds), Electioneering: A Comparative Study of Continuity
and Change. Oxford: Clarendon Press, pp. 244–77.
Bartolini, Stefano (2000) The Political Mobilization of the European Left, 1860–1980:
The Class Cleavage. Cambridge: Cambridge University Press.
Bartolini, Stefano and Peter Mair (1990) Identity, Competition and Electoral Availability:
The Stabilisation of European Electorates 1885–1985. Cambridge: Cambridge
University Press.
Barzelay, Michael (1992) Breaking through Bureaucracy: A New Vision for Managing
in Government. Berkeley, CA: University of California Press.
Bawn, Kathleen, Martin Cohen, David Karol et al. (2012) “A Theory of Political
Parties: Groups, Policy Demands and Nominations in American Politics.” Perspec-
tives on Politics 10 (3): 571–97.
Beaumont, Peter (2011) “It’s not just our leaders who are in a crisis. Democracy itself is
failing.” Guardian, November 20, http://www.theguardian.com/commentisfree/
2011/nov/20/peter-beaumont-democracy-in-crisis.
Beer, Samuel (1969) British Politics in the Collectivist Age. New York: Random House.
Bell, Joshua (2014) “ ‘The Tory Party at Prayer’: An Assessment of the Anglican
Voters’ Association with the Conservative Party.” Diploma thesis, University of
Lincoln.
Bellamy, Richard and Uta Staiger (eds) (2013) The Eurozone Crisis and the Democratic
Deficit. London: UCL European Institute.
Benoit, Kenneth (2004) “Models of Electoral System Change.” Electoral Studies 23
(3, September): 363–89.
Benoit, Kenneth and Michael Laver (2006) Party Policy in Modern Democracies. New
York: Routledge.
Berelson, Bernard R., Paul F. Lazarsfeld, and William N. McPhee (1954) Voting:
A Study of Opinion Formation in a Presidential Campaign. Chicago, IL: University
of Chicago Press.
Best, Heinrich and Maurizio Cotta (eds) (2000) Parliamentary Representatives in
Europe 1848–2000: Legislative Recruitment and Careers in Eleven European Coun-
tries. Oxford: Oxford University Press.
Beus, Jos de (2011) “Audience Democracy: An Emerging Pattern in Postmodern
Political Communication,” in Kees Brants and Katrin Voltmer (eds), Political
Communication in Postmodern Democracy: Challenging the Primacy of Politics.
Houndmills: Palgrave Macmillan.
References 191
Beyme, Klaus van (1996) “The Concept of Political Class: A New Dimension of
Research on Elites?” West European Politics 19: 68–87.
Bickerton, Christopher and Carlo Invernizzi Accetti (2017) “Populism and Technocracy:
Opposites or Complements?” Critical Review of International Social and Political
Philosophy 20 (2): 186–206.
Biezen, Ingrid van (1998) “Building Party Organisations and the Relevance of Past
Models: The Communist and Socialist Parties in Spain and Portugal.” West Euro-
pean Politics 21 (2): 32–62.
Biezen, Ingrid van (2003) Political Parties in New Democracies: Party Organization in
Southern and East-Central Europe. Basingstoke: Palgrave Macmillan.
Biezen, Ingrid van (2004) “Political Parties as Public Utilities.” Party Politics 10 (6):
701–22.
Biezen, Ingrid van (2009) “Constitutionalizing Party Democracy: The Constitutive
Codification of Political Parties in Post-War Europe.” Working paper series on the
legal regulation of political parties, no. 3, November.
Biezen, Ingrid van and Petr Kopecký (2007) “The State and the Parties: Public
Funding, Public Regulation and Rent-Seeking in Contemporary Democracies.”
Party Politics 13 (2): 235–54.
Biezen, Ingrid van and Petr Kopecký (2017) “The Paradox of Party Funding: The
Limited Impact of State Subsidies on Party Membership,” in Susan Scarrow, Paul
Webb, and Thomas Poguntke (eds), Organizing Political Parties: Representation,
Participation, and Power. Oxford: Oxford University Press.
Biezen, Ingrid van and Karl-Heinz Nassmacher (2001) “Political Finance in Southern
Europe: Italy, Portugal, and Spain,” in Karl-Heinz Nassmacher (ed.), Foundations
for Democracy: Approaches to Comparative Political Finance. Baden-Baden:
Nomos, pp. 131–54.
Biezen, Ingrid van and Michael Saward (2008) “Democratic Theorists and Party
Scholars: Why They Don’t Talk to Each Other and Why They Should.” Perspectives
on Politics 6 (1): 21–35.
Biezen, Ingrid van, Peter Mair, and Thomas Poguntke (2012) “Going, Going . . .
Gone? The Decline of Party Membership in Contemporary Europe.” European
Journal of Political Research 51 (1): 24–56.
Bille, Lars (2001) “Democratizing and Democratic Procedure: Myth or Reality.”
Bischoff, Carina (2005) “Political Competition and Contestability: A Study of Barriers
to Entry in 21 Democracies.” PhD dissertation, Department of Political and Social
Sciences, European University Institute.
Blechinger, Verena and Karl-Heinz Nassmacher (2001) “Political Finance in Non-
Western Democracies: Japan and Israel,” in Karl-Heinz Nassmacher (ed.), Founda-
tions for Democracy: Approaches to Comparative Political Finance. Baden-Baden:
Nomos, pp. 155–80.
Blumler, Jay and Stephen Coleman (2010) “Political Communication in Freefall: The
British Case—and Others?” International Journal of Press/Politics 15 (2): 139–54.
Blyth, Mark (2002) Great Transformations: Economic Ideas and Institutional Change in
the Twentieth Century. Cambridge: Cambridge University Press.
192 References
Blyth, Mark and Richard S. Katz (2005) “From Catch-All Politics to Cartelisation.”
West European Politics 28 (January): 33–60.
Boda, Michael (2004) Revisiting Free and Fair Elections. Geneva: Inter-Parliamentary
Union.
Bolleyer, Nicole (2009) “Inside the Cartel Party.” Political Studies 57 (3): 559–79.
Bolleyer, Nicole (2011) “The Influence of Political Parties on Policy Coordination.”
Governance 24 (3): 469–94.
Borchert, Jens (2000) “The Political Class and Its Self-Interested Theory of Democ-
racy: Historical Developments and Institutional Consequences.” European Consor-
tium for Political Research Joint Sessions of Workshops, Copenhagen.
Borchert, Jens and Jürgen Zeiss (eds) (2003) The Political Class in Advanced Democ-
racies: A Comparative Handbook. Oxford: Oxford University Press.
Bottom, Karin (2007) “Party Cartelisation and the Locality: The British Case.” Paper
presented at ECPR Joint Sessions of Workshops, Helsinki, May 7–12.
Bowler, Shaun, David Farrell, and Richard Katz (eds) (1999) Party Discipline and
Parliamentary Government. Columbus, OH: Ohio State University Press.
Bowler, Shaun, Elisabeth Carter, and David M. Farrell (2001) “Studying Electoral
Institutions and Their Consequences: Electoral Systems and Electoral Laws.” Uni-
versity of California, Irvine, Center for the Study of Democracy, paper 01-05. http://
repositories.cdlib.org/csd/01-05.
Brans, Marleen (2012) Rewards for High Public Office in Europe and North America.
Hoboken, NJ: Taylor and Francis.
Brans, Marleen and Annie Hondeghem (1999) The Senior Civil Service in Belgium.
Oxford: Oxford University Press.
Brants, Kees and Philip van Praag (2006) “Signs of Media Logic: Half a Century of
Political Communication in the Netherlands.” Javnost: The Public 13 (1): 25–30.
Brennan, Jason (2016) Against Democracy. Princeton, NJ: Princeton University Press.
Bruter, Michael and Sarah Harrison (2009) The Future of Our Democracies? Young
Party Members in 6 European Democracies. Houndmills: Palgrave Macmillan.
Buchanan, James and Gordon Tullock (1962) The Calculus of Consent: Logical
Foundations of Constitutional Democracy. Ann Arbor, MI: University of Michigan
Press.
Burke, Edmund (1770) “Thoughts on the cause of the present discontents.”
http://press-pubs.uchicago.edu/founders/documents/v1ch13s6.html.
Burke, Edmund (1774) “Speech to the electors of Bristol.” http://press-pubs.uchicago.
edu/founders/documents/v1ch13s7.html.
Butler, David and Austin Ranney (eds) (1992) Electioneering: A Comparative Study of
Continuity and Change. Oxford: Clarendon Press.
Butler, David and Donald Stokes (1969) Political Change in Britain. New York:
St. Martins Press.
Caramani, Daniele (2004) The Nationalization of Politics: The Formation of National
Electorates and Party Systems in Western Europe. Cambridge: Cambridge University
Press.
Caramani, Daniele (2006) “Is There a European Electorate and What Does It Look
Like? Evidence from Electoral Volatility Measures, 1976–2004.” West European
Politics 29 (1): 1–27.
References 193
Caramani, Daniele (2015) “Will versus reason: The populist and technocratic challenge
to party democracy.” Paper presented at the ECPR Joint Sessions of Workshops,
Warsaw, March 29–April 2.
Cartwright, Edward (2011) Behavioral Economics. Hoboken, NJ: Taylor and Francis.
Carty, R. Kenneth (2004) “Parties as Franchise Systems: The Stratarchical Organiza-
tional Imperative.” Party Politics 10 (1): 5–24.
Castles, Francis and Peter Mair (1984) “Left-Right Political Scales: Some ‘Expert’
Judgements.” European Journal of Political Research 12 (1): 73–88.
Castles, Francis and Rudolf Wildenmann (eds) (1986) Visions and Realities of Party
Government. Berlin: de Gruyter.
Christensen, Jrgen Grnnegaard (2006) “Ministers and Mandarins under Danish
Parliamentarism.” International Journal of Public Administration 29 (12):
997–1019.
Cicero, Quintus Tullius (2012 [64BC]) How to Win an Election: An Ancient Guide
for Modern Politicians, translated by Philip Freeman. Princeton, NJ: Princeton
University Press.
Clancy, Paula and Grainne Murphy (2006) “Outsourcing Government: Public Bodies
and Accountability.” Sandbach: TASC New Island.
Collier, David and Steven Levitsky (1997) “Democracy with Adjectives: Conceptual
Innovation in Comparative Research.” World Politics 49 (3): 430–51.
Collignon, Stefan (2015) “More democracy, or no more Europe.” http://europesworld.
org/2015/10/06/democracy-europe/#.WRG-08a1uUk.
Colomer, Josep (2005) “It’s Parties that Choose Electoral Systems (or, Duverger’s
Laws Upside Down).” Political Studies 53 (1): 1–21.
Commission d’enquête sur l’exploitation et la construction des chemins de fer (1858)
Enquête sur les Moyen d’Assurer la Regularité et la Surêté de l’Exploitation sur les
Chemins de Fer. Paris: Imprimerie Impériale.
Cotta, Maurizio and Heinrich Best (2000) “Between Professionalisation and Democ-
ratisation: A Synoptic View on the Making of the European Representative,” in
Heinrich Best and Maurizio Cotta (eds), Parliamentary Representatives in Europe
1848–2000: Legislative Recruitment and Careers in Eleven European Countries.
Crewe, Ivor (1974) “Do Butler and Stokes Really Explain Political Change in
Britain?” European Journal of Political Research 2: 47–92.
Cross, William P. (2015) “Party Membership in Canada,” in Emilie van Haute and
Anika Gauja (eds), Party Members and Activists. Hoboken, NJ: Taylor and Francis,
pp. 50–65.
Cross, William P. and Richard S. Katz (eds) (2013) The Challenges of Intra-Party
Democracy. Oxford: Oxford University Press.
Cross, William P., Ofer Kenig, Scott Pruysers, and Gideon Rahat (2016) The Promise
and Challenge of Party Primary Elections: A Comparative Perspective. Montreal:
McGill-Queen’s University Press.
Crossman, Richard (1972) The Myths of Cabinet Government. Cambridge, MA:
Harvard University Press.
Crozier, M., Samuel Huntington, and Joji Watanuki (1975) The Crisis of Democracy.
New York: New York University Press.
194 References
D’Aspremont, C., R. Dos Santos Ferreira, L.-A. Gérard-Varet (1991) “Pricing
Schemes and Cournotian Equilibria.” American Economic Review 81 (3): 666–73.
Dahl, Robert (1956) A Preface to Democratic Theory. Chicago, IL: University of
Chicago Press.
Dahl, Robert (1961) Who Governs? Democracy and Power in an American City. New
Haven, CT: Yale University Press.
Dahl, Robert (1971) Polyarchy: Participation and Opposition. New Haven, CT: Yale
University Press.
Dalton, Russell (1984) “Cognitive Mobilization and Partisan Dealignment in
Advanced Industrial Democracies.” Journal of Politics 46 (1): 264–84.
Dalton, Russell (1988) Citizen Politics in Western Democracies: Public Opinion and
Political Parties in the United States, Great Britain, West Germany, and France.
Chatham, NJ: Chatham House.
Dalton, Russell (2008) Citizen Politics: Public Opinion and Political Parties in
Advanced Industrial Democracies (5th ed.). Washington, DC: CQ Press.
Dalton, Russell J. and Martin P. Wattenberg (eds) (2000) Parties without Partisans:
Political Change in Advanced Industrial Democracies. Oxford: Oxford University
Press.
Dathan, Matt (2015) “So, who are the ‘moronic MPs’ who nominated Jeremy Corbyn
for the Labour leadership contest?” http://www.independent.co.uk/news/uk/pol
itics/who-are-the-morons-who-nominated-jeremy-corbyn-for-the-labour-leadership-
contest-10406527.html, accessed September 24, 2015.
Depauw, Sam and Shane Martin (2009) “Legislative Party Discipline and Cohesion in
Comparative Perspective,” in Daniela Giannetti and Kenneth Benoit (eds), Intra-
Party Politics and Coalition Governments. London: Routledge, pp. 103–20.
Deschouwer, Kris (1994) “The Decline of Consociationalism and the Reluctant Mod-
ernization of Belgian Mass Parties,” in Richard S. Katz and Peter Mair (eds), How
Parties Organize: Change and Adaptation in Party Organizations in Western Dem-
ocracies. London: Sage, pp. 80–108.
Detterbeck, Klaus (2016) “Candidate Selection in Germany: Local and Regional Party
Elites Still in Control?” American Behavioral Scientist 60 (7): 837–52.
Detterbeck, Klaus and Wolfgang Renzsch (2003) “Multi-Level Electoral Competition:
The German Case.” European Urban and Regional Studies 10 (3): 257–69.
Diamond, Peter and Hannu Vartiainen (eds) (2007) Behavioral Economics and Its
Applications. Princeton, NJ: Princeton University Press.
Dőring, Herbert (1987) “Party Government in Britain: Recent Conspicuous Con-
straints,” in Richard S. Katz (ed.), Party Governments: European and American
Experiences. Berlin: de Gruyter, pp. 118–54.
Dogan, Mattei (2001) “Class, Religion, Party: Triple Decline of Electoral Cleavages in
Western Europe,” in L. Karvonen and S. Kuhnle (eds), Party Systems and Voter
Alignments Revisited. London: Routledge.
Dogan, Mattei (2002) “Accelerated Decline of Religious Beliefs in Europe.” Com-
parative Sociology 1 (2): 127–49.
Downs, Anthony (1957) An Economic Theory of Democracy. New York: Harper
and Row.
References 195
Duverger, Maurice (1959 [1951]) Political Parties: Their Organization and Activity in
the Modern State. New York: John Wiley and Sons.
Duverger, Maurice (1986) “Duverger’s Law: Forty Years Later,” in Bernard Grofman
and Arend Lijphart (eds), Electoral Laws and Their Political Consequences.
New York: Agathon Press, pp.69–84.
Easton, David (1975) “Re-Assessment of the Concept of Political Support.” British
Journal of Political Science 5 (4): 435–57.
EB81 Report (2014) “Pubic Opinion in the European Union.” http://ec.europa.eu/
commfrontoffice/publicopinion/archives/eb/eb81/eb81_publ_en.pdf.
Eddy, Melissa (2017) “Merkel puts restrictions on refugees.” New York Times, October
10, p. A8.
Egeberg, M. (1998) “Causes of bureaucratic autonomy: The impact of organizational
and cultural factors.” Working Paper 98/15. Centre for European Studies, Univer-
sity of Oslo.
Eggers, Andrew and Jens Hainmueller (2009) “MPs for Sale? Returns to Office in
Postwar British Politics.” American Political Science Review 103 (4): 513–33.
Eldersveld, Samuel (1964) Political Parties: A Behavioral Analysis. Chicago, IL: Rand
McNally.
Epstein, Leon (1967) Political Parties in Western Democracies. New York: Praeger.
Epstein, Leon (1986) Political Parties in the American Mold. Madison, WI: University
of Wisconsin Press.
Fabbrini, Sergio and Marc Lazar (2016) “La leadership di Renzi tra partito e gov-
erno,” in Maurizio Carbone and Simona Piattoni (eds), Politica in Italia. I fatti
dell’anno e le interpretazioni. Edizione 2016. Bologna: Il Mulino.
Farneti, Paolo (1985) The Italian Party System (1945–1980). London: Frances Pinter.
Fenno, Richard F., Jr. (1978) Home Style: House Members in Their Districts. New
York: HarperCollins.
Finer, Samuel E. (1970) Comparative Government: An Introduction to the Study of
Politics. Harmondsworth: Penguin.
Finer, Samuel E. (1975) Adversary Politics and Electoral Reform. London: Anthony Wigram.
Fisher, Justin, Edward Fieldhouse, and David Cutts (2014) “Members Are Not the
Only Fruit: Volunteer Activity in British Political Parties at the 2010 General
Election.” British Journal of Politics and International Relations 16 (1): 75–95.
Flinders, Matthew and Felicity Matthews (2012) “Party Patronage in the United
Kingdom: A Pendulum of Public Appointments,” in Petr Kopecký, Peter Mair,
and Maria Spirova (eds), Party Patronage and Party Government in European
Democracies. Oxford: Oxford University Press, pp. 335–56.
Foxe, Ken (2011) Snouts in the Trough: Irish Politicians and Their Expenses. Dublin:
Gill and Macmillan.
Freedland, Jonathan (2016) “Welcome to the age of Trump.” Guardian, May 19.
https://www.theguardian.com/us-news/2016/may/19/welcome-to-the-age-of-trump?
utm_content=buffer99672&utm_medium=social&utm_source=twitter.com&utm_
campaign=buffer.
Galbraith, John Kenneth (1992) The Culture of Contentment. Boston, MA: Houghton
Mifflin.
196 References
Gallagher, Michael, Michael Laver, and Peter Mair (2011) Representative Government
in Modern Europe (5th ed.). Maidenhead: McGraw-Hill Education.
Galli, Giorgio and Alfanso Prandi (1970) Patterns of Political Participation in Italy.
New Haven, CT: Yale University Press.
Gauja, Anika (2013) The Politics of Party Policy: From Members to Legislators.
Basingstoke: Palgrave Macmillan.
Gauja, Anika (2015) “Agitating for change? The role of Australian party members in
party reform.” Paper presented at the General Conference of the European Con-
sortium for Political Research, Montreal, August 27–29.
Ginsberg, Benjamin (1982) The Consequences of Consent: Elections, Citizen Control
and Popular Acquiescence. New York: Random House.
Goodwin-Gill, Guy (1994) Free and Fair Elections. Geneva: Inter-Parliamentary
Union.
Goodwin-Gill, Guy (2006) Free and Fair Elections: New Expanded Edition. Geneva:
Inter-Parliamentary Union.
Haas, Peter (1992) “Introduction: Epistemic Communities and International Policy
Coordination.” International Organization 46 (1): 1–35.
Hagevi, Magnus (2003) “Sweden: Between Participation Ideal and Professionalism,”
in Jens Borchert and Jürgen Zeiss (eds), The Political Class in Advanced Democra-
cies: A Comparative Handbook. Oxford: Oxford University Press, pp. 352–73.
Halman, L. and O. Riis (2003) “Contemporary European Discourses on Religion and
Morality,” in L. Halman and O. Riis (eds), Religion in Secularizing Society: The
Europeans’ Religion at the End of the 20th Century. Leiden: Brill, pp. 1–21.
Halman, L., Ronald Inglehart, Jaime Díez-Medrano et al. (2008) Changing Values and
Beliefs in 85 Countries: Trends from the Values Surveys from 1981 to 2004. Leiden:
Brill.
Hazan, Reuven Y. and Gideon Rahat (2010) Democracy within Parties: Candidate
Selection Methods and Their Political Consequences. Oxford: Oxford University
Press.
Head, Brian (2010) “Reconsidering Evidence-Based Policy: Key Issues and Chal-
lenges.” Policy and Society 29 (2): 77–94.
Heidar, Knut and Ruud Koole (2000) “Parliamentary Party Groups Compared,” in
Knut Heidar and Ruud Koole (eds), Parliamentary Party Groups in European
Democracies. London: Routledge, pp. 248–70.
Heller, William and Carol Mershon (2005) “Party Switching in the Italian Chamber of
Deputies, 1996–2001.” Journal of Politics 67 (2): 536–59.
Herman, Valentine and Françoise Mendel (1976) Parliaments of the World: A Reference
Compendium. Berlin: De Gruyter.
Hirschman, Albert (1970) Exit, Voice, and Loyalty: Responses to Decline in Firms,
Organizations, and States. Cambridge, MA: Harvard University Press.
Hood, Christopher (1991) “A Public Management for all Seasons?” Public Adminis-
tration 69: 3–19.
Hooghe, Marc and Anna Kern (2015) “Party Membership and Closeness and the
Development of Trust in Political Institutions: An Analysis of the European Social
Survey, 2002–2010.” Party Politics 21 (6): 944–56.
Hopkin, Jonathan and Caterina Paolucci (1999) “New Parties and the Business Firm
Model of Party Organization: Cases from Spain and Italy.” European Journal of
Political Research 35 (3): 307–39.
References 197
Hotelling, Harold (1929) “Stability in Competition.” Economic Journal 39 (153):
41–57.
Houska, Joseph (1985) Influencing Mass Political Behavior: Elites and Political Sub-
cultures in the Netherlands and Austria. Berkeley, CA: Institute of International
Studies, University of California.
Huber, John and Ronald Inglehart (1995) “Expert Interpretations of Party Space and
Party Locations in 42 Societies.” Party Politics 1 (1): 73–111.
Huntington, Samuel (1966) The Clash of Civilizations and the Remaking of World
Order. New York: Simon and Schuster.
Ignazi, Piero (2014) “Power and the (Il)legitimacy of Political Parties: An Unavoidable
Paradox of Contemporary Democracy?” Party Politics 20 (2): 160–9.
Implement Record (1922) “Jesting and the Constitutions.” Implement Record: Tractors
and Farm Equipment 19 (11): 34.
Independent Parliamentary Standards Authority (2012) “Reviewing MPs’ pay and
pensions: A consultation.” http://parliamentarystandards.org.uk/payandpensions/
Documents/02.%20Reviewing%20MPs%27%20Pay%20and%20Pensions%20-%
20A%20Consultation.pdf.
Inglehart, Ronald (1970) “Cognitive Mobilization and European Identity.” Compara-
tive Politics 3 (1): 45–70.
Inglehart, Ronald (1990) Culture Shift in Advanced Industry Society. Princeton, NJ:
Princeton University Press.
Janda, Kenneth (2009) “Laws against party switching, defecting, or floor-crossing in
national parliaments.” Prepared for delivery at the 2009 World Congress of the
International Political Science Association Santiago, Chile, July 12–16.
Jay, Antony and Jonathan Lynn (1988) “Power to the People.” Yes, Prime Minister,
BBC, series 2, episode 5.
Jennings, Ivor (1969) Cabinet Government. Cambridge: Cambridge University
Press.
Jensen, Hanne Nex and Tim Knudsen (1999) “Senior Officials in the Danish Central
Administration: From Bureaucrats to Policy Professionals and Managers,” in
Edward Page and Vincent Wright (eds), Bureaucratic Élites in Western European
States. Oxford: Oxford University Press, pp. 229–48.
Jensen, Torben K. (2003) “Denmark: Professionalism in an Egalitarian Political Culture,”
in Jens Borchert and Jürgen Zeiss (eds), The Political Class in Advanced Democracies:
A Comparative Handbook. Oxford: Oxford University Press, pp. 84–106.
Jouvenel, Robert de (1914) La Republique des Camarades. Paris: Bernard Grasset.
Jun, Uwe (2003) “Great Britain: From the Prevalence of the Amateur to the Domin-
ance of the Professional Politician,” in Jens Borchert and Jürgen Zeiss (eds), The
Political Class in Advanced Democracies: A Comparative Handbook. Oxford:
Kaid, Lynda Lee and Jesper Strömbäck (2008) “Election News Coverage around the
World: A Comparative Perspective,” in Jesper Strömbäck and Lynda Lee Kaid
(eds), The Handbook of Election News Coverage around the World. New York:
Routledge.
Kaiser, André (1997) “Types of Democracy: From Classical to New Institutionalism.”
Journal of Theoretical Politics 9 (4): 419–44.
Kalyvas, Stathis (1996) The Rise of Christian Democracy in Europe. Ithaca, NY:
Cornell University Press.
198 References
Kalyvas, Stathis and Kees van Kersbergen (2010) “Christian Democracy.” Annual
Review of Political Science 13: 183–209.
Karvonen, Lauri (2007) “Legislation on Political Parties: A Global Comparison.”
Katz, Richard S. (1986) “Party Government: A Rationalistic Conception,” in
G. Castles and R. Wildenmann (eds), Visions and Realities of Party Government.
Berlin: de Gruyter.
Katz, Richard S. (1987) “Party Government and Its Alternatives,” in Richard
S. Katz (ed.), Party Governments: European and American Experiences. Berlin: de
Gruyter.
Katz, Richard S. (1997) Democracy and Elections. Oxford: Oxford University Press.
Katz, Richard S. (2002) “The Internal Life of Parties,” in Kurt Richard Luther and
Ferdinand Müller-Rommel (eds), Political Parties in the New Europe. Oxford:
Katz, Richard S. (2003) “Cartels, consensus democracy, and representation in the
European Union.” Paper presented at the European Consortium for Political
Research Joint Sessions of Workshops, Edinburgh.
Katz, Richard S. (2005) “Why Are There so Many (or so Few) Electoral Reforms?” in
Michael Gallagher and Paul Mitchell (eds), The Politics of Electoral Systems.
Katz, Richard S. (2013) “Should We Believe that Improved Intra-Party Democracy
Would Arrest Party Decline?” in William P. Cross and Richard S. Katz (eds), The
Challenges of Intra-Party Democracy. Oxford: Oxford University Press, pp. 49–64.
Katz, Richard S. (2014) “No Man Can Serve Two Masters: Party Politicians, Party
Members, Citizens, and Principal-Agent Models of Democracy.” Party Politics 20
(2, March): 183–93.
Katz, Richard S. (2017) “Afterword,” in Susan Scarrow, Paul Webb, and Thomas
Poguntke (eds), Organizing Political Parties: Representation, Participation, and
Power. Oxford: Oxford University Press.
Katz, Richard S. and Robin Kolodny (1994) “Party Organization as an Empty Vessel:
Parties in American Politics,” in Richard S. Katz and Peter Mair (eds), How Parties
Organize: Adaptation and Change in Party Organizations in Western Democracies.
London: Sage.
Katz, Richard S. and Peter Mair (1992a) “Changing Models of Party Organization:
The Emergence of the Cartel Party.” Workshop on Democracy and the Organiza-
tion of Political Parties, European Consortium for Political Research Joint Sessions
of Workshops, University of Limerick, Ireland.
Katz, Richard S. and Peter Mair (1992b) “Introduction,” in Richard S. Katz and Peter
Mair (eds), Party Organizations: A Data Handbook. London: Sage.
Katz, Richard S. and Peter Mair (eds) (1992c) Party Organizations: A Data Handbook.
London: Sage.
Katz, Richard S. and Peter Mair (1993) “The Evolution of Party Organizations in
Europe: The Three Faces of Party Organization.” American Review of Politics 14
(Winter): 593–618.
Katz, Richard S. and Peter Mair (eds) (1994) How Parties Organize: Change and
Adaptation in Party Organizations in Western Democracies. London: Sage.
References 199
Katz, Richard S. and Peter Mair (1995) “Changing Models of Party Organization and
Party Democracy: The Emergence of the Cartel Party.” Party Politics 1: 5–28.
Katz, Richard S. and Peter Mair (1996) “Cadre, Catch-All or Cartel? A Rejoinder.”
Katz, Richard S. and Peter Mair (2002) “The Ascendancy of the Party in Public Office:
Party Organizational Change in Twentieth-Century Democracies,” in Richard
Gunther, José Ramón Montero, and Juan J. Linz (eds), Political Parties: Old
Concepts and New Challenges. Oxford: Oxford University Press, pp. 113–35.
Katz, Richard S. and Peter Mair (2009) “The Cartel Party Thesis Revisited.” Perspectives
on Politics 7 (4): 753–66.
Katz, Richard S., Peter Mair, Luciano Bardi et al. (1992) “The Membership of
Political Parties in European Democracies, 1960–1990.” European Journal of Polit-
ical Research 22 (3): 329–45.
Kennedy, John F. (1957) Profiles in Courage. New York: Harper and Brothers.
Key, V. O., Jr. (1964) Politics, Parties and Pressure Groups. New York: Crowell.
Kirchheimer, Otto (1957) “The Waning of Opposition in Parliamentary Regimes.”
Social Research 24: 127–56.
Kirchheimer, Otto (1966) “The Transformation of Western European Party Systems,”
in Joseph LaPalombara and Myron Weiner (eds), Political Parties and Political
Development. Princeton, NJ: Princeton University Press, pp. 177–200.
Kitschelt, Herbert (2000) “Citizens, Politicians, and Party Cartellization: Political
Representation and State Failure in Post-Industrial Democracies.” European Jour-
nal of Political Research 37 (2): 149–79.
Koole, Ruud (1994) “The Vulnerability of the Modern Cadre Party in the Netherlands,”
in Richard S. Katz and Peter Mair (eds), How Parties Organize: Change and Adapta-
tion in Party Organizations in Western Democracies. London: Sage, pp. 278–303.
Koole, Ruud (1996) “Cadre, Catch-All or Cartel? A Comment on the Notion of the
Cartel Party.” Party Politics 2 (4): 507–23.
Koole, Ruud and J. J. M. van Holsteyn (1999) “Partijleden in perspectief: Kiezers en
leden vergeleken,” in Jaarboek Documentatiecentrum Nederlandse Politieke Partijen
1999. Groningen: DNPP, pp. 93–124.
Koole, Ruud and Philip van Praag (1990) “Electoral Competition in a Segmented
Society: Campaign Strategies and the Importance of Elite Perceptions.” European
Journal of Political Research 18 (1): 51–69.
Kopecký, Petr (1995) “Developing Party Organizations in East-Central Europe: What
Type of Party Is Likely to Emerge?” Party Politics 1 (4): 515–34.
Kopecký, Petr and Peter Mair (2012) “Party Patronage as an Organizational Resource,”
in Petr Kopecký, Peter Mair, and Maria Spirova (eds), Party Patronage and Party
Government in European Democracies. Oxford: Oxford University Press, pp. 2–38.
Kopecký, Petr and G. Scherlis (2008) “Party Patronage in Contemporary Europe.”
European Review 16 (3): 355–71.
Kosiara-Pedersen, Karina, Susan Scarrow, and Emelie van Haute (2017) “Rules of
Engagement? Party Membership Costs, New Forms of Party Affiliation, and Par-
tisan Participation,” in Susan Scarrow, Paul Webb, and Thomas Poguntke (eds),
Organizing Political Parties: Representation, Participation, and Power. Oxford:
200 References
Kriesi, Hanspeter, Edgar Grande, Romain Lachat et al. (2008) West European Politics
in the Age of Globalization. Cambridge: Cambridge University Press.
Krouwel, André (2003) “Otto Kirchheimer and the Catch-all Party.” West European
Politics 26 (2): 23–40.
Krouwel, André (2006) “Party Models,” in Richard S. Katz and William Crotty (eds),
Handbook of Party Politics. London: Sage, pp. 249–69.
Laakso, Markku and Rein Taagepera (1979) “ ‘Effective’ Number of Parties: A Measure
with Application to West Europe.” Comparative Political Studies 12 (1): 3–27.
Laclau, Ernesto (2005) On Populist Reason. London: Verso.
LaPalombara, Joseph (1964) Interest Groups in Italian Politics. Princeton, NJ: Princeton
University Press.
Laver, Michael and Ian Budge (1992) “Measuring Policy Distances and Modelling
Coalition Formation,” in Michael Laver and Ian Budge (eds), Party Policy and
Government Coalitions. Basingstoke: Macmillan, pp. 15–40.
Law, C. M. (1967) “The Growth of Urban Population in England and Wales,
1801–1911.” Transactions, Institute of British Geographers 41: 125–43.
Lawson, Kay (ed.) (1980) Political Parties and Linkage: A Comparative Perspective.
New Haven, CT: Yale University Press.
Lawson, Kay (2011) “Party Linkage,” in Bertrand Badie, Dirk Berg-Schlosser, and
Leonardo Morlino (eds), International Encyclopedia of Political Science. Thousand
Oaks, CA: Sage Publications, pp. 1812–14.
Lebaron, Frédéric (2000) “The Space of Economic Neutrality: Types of Legitimacy
and Trajectories of Central Bank Managers.” International Journal of Contempor-
ary Sociology 37 (2): 208–29.
Levinson, Marc (2016) An Extraordinary Time: The End of the Postwar Boom and the
Return of the Ordinary Economy. New York: Basic Books.
Lewis-Beck, Michael (1990) Economics and Elections: The Major Western Democracies.
Ann Arbor, MI: University of Michigan Press.
Liegel, Barbara and Wolfgang Müller (1999) “Senior Officials in Austria,” in Edward
Page and Vincent Wright (eds), Bureaucratic Élites in Western European States.
Oxford: Oxford University Press, pp. 90–120.
Lijphart, Arend (1968) The Politics of Accommodation: Pluralism and Democracy in
the Netherlands. Berkeley, CA: University of California Press.
Lijphart, Arend (1999) Patterns of Democracy. New Haven, CT: Yale University Press.
Lindberg, Leon and Stuart Scheingold (1970) Europe’s Would-Be Polity: Patterns of
Change in the European Community. Englewood Cliffs, NJ: Prentice-Hall.
Linz, Juan and Alfred Stepan (1996) “Toward Consolidated Democracies.” Journal of
Democracy 7 (2): 14–33.
Lipinski, Daniel, William T. Bianco, and Ryan Work (2003) “What Happens When
House Members ‘Run with Congress’? The Electoral Consequences of Institutional
Loyalty.” Legislative Studies Quarterly 28 (3): 413–29.
Lipset, Seymour Martin (1960) Political Man: The Social Bases of Politics. New York:
Doubleday.
Lipset, Seymour Martin and Stein Rokkan (1967) “Cleavage Structures, Party Systems
and Voter Alignments: An Introduction,” in S. M. Lipset and S. Rokkan (eds),
Party Systems and Voter Alignments. New York: Free Press, pp. 1–64.
References 201
Long, Larry (1991) “Residential Mobility Differences among Developed Countries.”
International Regional Science Review 14 (2): 133–47.
Lorwin, Val (1971) “Segmented Pluralism: Ideological Cleavages and Political Cohe-
sion in the Smaller European Democracies.” Comparative Politics 3 (2): 141–75.
Mac Con Uladh, Damian (2013) “Greece: Democracy under threat in the turmoil.”
Irish Times, September 12. https://www.irishtimes.com/business/economy/greece-
democracy-under-threat-in-the-turmoil-1.1524023.
Madison, James (1961 [1787]) “The Federalist No. 10,” in The Federalist Papers.
New York: Penguin Books.
Mair, Peter (1993) “Myths of Electoral Change and the Survival of Traditional Parties:
The 1992 Stein Rokkan Lecture.” European Journal of Political Research 24 (2):
121–33.
Mair, Peter (1994) “Party Organizations: From Civil Society to the State,” in Richard
S. Katz and Peter Mair (eds), How Parties Organize: Change and Adaptation in
Party Organizations in Western Democracies. London: Sage, pp. 1–22.
Mair, Peter (2009) “Representative versus responsible government.” Max-Planck-
Institut für Gesellschaftsforschung working paper 09/8.
Mair, Peter (2013) Ruling the Void: The Hollowing of Western Democracy. London:
Verso.
Majone, Giandomenico (1994) “The Rise of the Regulatory State in Europe.” West
European Politics 17 (3): 77–101.
Majone, Giandomenico (1997) “From the Positive to the Regulatory State: The
Causes and Consequences of Changes in the Mode of Governance.” Journal of
Public Policy 17 (2): 139–67.
Manin, Bernard (1997) The Principles of Representative Government. Cambridge:
Cambridge University Press.
Marx, Karl (1852) “The Eighteenth Brumaire of Louis Bonaparte.” https://www.
marxists.org/archive/marx/works/1852/18th-brumaire/ch01.htm.
Masket, Seth E. (2009) No Middle Ground: How Informal Party Organizations Control
Nominations and Polarize Legislatures. Ann Arbor, MI: University of Michigan Press.
May, John D. (1973) “Opinion Structure of Political Parties: The Special Law of
Curvilinear Disparity.” Political Studies 21 (2): 135–51.
McClosky, Herbert (1964) “Consensus and Ideology in American Politics.” American
Political Science Review 58 (2): 361–82.
McKenzie, Robert (1963) British Political Parties: The Distribution of Power within the
Conservative and Labour Parties. New York: Praeger.
McKibbin, Ross (2010) Parties and People: England 1914–1951. Oxford: Oxford
University Press.
McRae, Kenneth (1974) Consociational Democracy: Political Accommodation in
Segmented Societies. Toronto: McClelland and Stewart.
Michels, Robert (1962 [1911]) Political Parties: A Sociological Study of the Oligarchic
Tendencies of Modern Democracy. New York: Collier Books.
Mill, John Stuart (1982 [1861]) Considerations on Representative Government. Chicago,
IL: Henry Regnery.
Miller, Gary J. (2005) “The Political Evolution of Principal-Agent Models.” Annual
Review of Political Science 8: 203–25.
202 References
Minkin, L. (1978) The Labour Party Conference. London: Allen Lane.
Mitchell, G. J. (1997) Not for America Alone: The Triumph of Democracy and the Fall
of Communism. New York: Kodansha Amer.
Mosse, B. and E. Whitley (2009) “Critically Classifying UK E-Government Website
Benchmarking and the Recasting of the Citizen as Customer.” Information Systems
Journal 19 (2): 149–73.
Mudde, Cas (2010) “The Populist Radical Right: A Pathological Normalcy.” West
European Politics 33 (6): 1167–86.
Mudde, Cas (2013) “Three Decades of Populist Radical Right parties in Western
Europe: So What?” European Journal of Political Research 52: 1–19.
Mueller, Dennis C. (1989) Public Choice II. Cambridge: Cambridge University Press.
Müller, Wolfgang (1989) “Party Patronage in Austria: Theoretical Considerations and
Empirical Findings,” in Anton Pelinka and Fritz Plasser (eds), The Austrian Party
System. Boulder, CO: Westview Press, pp. 327–56.
Müller, Wolfgang (1993) “After the ‘Golden Age’: Research into Austrian
Political Parties since the 1980s.” European Journal of Political Research 23 (4):
439–63.
Müller, Wolfgang (1994) “The Development of Austrian Party Organizations in the
Post-War Period,” in Richard S. Katz and Peter Mair (eds), How Parties Organize:
Change and Adaptation in Party Organizations in Western Democracies. London:
Sage, pp. 51–79.
Müller, Wolfgang (2000) “Political Parties in Parliamentary Democracies: Making
Delegation and Accountability Work.” European Journal of Political Research 37:
335–51.
Müller, Wolfgang and Ulrich Sieberer (2005) “Party Law,” in Richard S. Katz and
William Crotty (eds), Handbook of Party Politics. London: Sage, pp. 435–45.
Müller, Wolfgang and Kaare Strm (eds) (1999) Policy, Office, or Votes? How Political
Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University
Press.
Mylonas, Harris (2016) “Greece.” European Journal of Political Research Political
Data Yearbook 55: 113–23.
Namier, Lewis (1963) The Structure of Politics at the Accession of George III. London:
Macmillan.
Narud, Hanne Marthe (2003) “Norway: Professionalization: Party-Oriented and
Constituency-Based,” in Jens Borchert and Jürgen Zeiss (eds), The Political Class
in Advanced Democracies: A Comparative Handbook. Oxford: Oxford University
Press, pp. 298–319.
Nassmacher, Karl-Heinz (ed.) (2001) Foundations for Democracy: Approaches to
Comparative Political Finance. Baden-Baden: Nomos.
Nassmacher, Karl-Heinz (2009) The Funding of Party Competition: Political Finance
in 25 Democracies. Baden-Baden: Nomos.
Neumann, Sigmund (1956) “Toward a Comparative Study of Political Parties,” in
Sigmund Neumann (ed.), Modern Political Parties. Chicago, IL: University of
Chicago Press, pp. 395–424.
Noble, T. (2000) “The Mobility Transition: Social Mobility Trends in the First Half of
the Twenty-First Century.” Sociology 34 (1): 35–51.
References 203
Nordlinger, Eric (1967) The Working-Class Tories: Authority, Deference and Stable
Democracy. Berkeley, CA: University of California Press.
Norris, Pippa (1997) “Political Communications.” Developments in British Politics 5:
75–88.
Nossiter, Adam (2017) “Disrespect intended: Mechanic-candidate bursts French pol-
itical elite’s bubble.” New York Times, April 6. https://www.nytimes.com/2017/04/
06/world/europe/with-disrespect-mechanic-candidate-bursts-french-political-elites-
bubble.html.
O’Donnell, James (1990) Word Gloss. Dublin: Institute of Public Administration.
O’Malley, Eoin, Stephen Quinlan and Peter Mair (2012) “Party Patronage in Ireland:
Changing Parameters,” in Petr Kopecký, Peter Mair, and Maria Spirova (eds),
Party Patronage and Party Government in European Democracies. Oxford: Oxford
University Press, pp. 206–29.
Olson, Mancur (1965) The Logic of Collective Action: Public Goods and the Theory of
Groups. Cambridge, MA: Harvard University Press.
Orth, Samuel (1919) The Boss and the Machine. New Haven, CT: Yale University Press.
Osborne, David and Ted Gaebler (1992) Reinventing Government: How the Entrepre-
neurial Spirit Is Transforming the Public Sector. New York: Addison-Wesley.
Oscarsson, Henrik and Sören Holmberg (2010) Swedish Voting Behavior. Gothenburg:
Swedish Election Studies Program, University of Gothenburg.
Ostrogorski, Moise (1903) La Démocratie et L’organisation des Partis Politique. Paris:
Calmann-Levy.
Pajala, Antti (2013) “Government vs Opposition Voting in the Finnish Parliament
Eduskunta since World War II.” European Journal of Government and Economics
2 (1): 41–58.
Panebianco, Angelo (1988) Political Parties: Organization and Power. Cambridge:
Cambridge University Press.
Pappalardo, Adriano (1981) “The Conditions for Consociational Democracy:
A Logical and Empirical Critique.” European Journal of Political Research 9:
365–90.
Parsons, Talcott (ed.) (1964) Max Weber: The Theory of Social and Economic Organ-
ization. New York: Free Press.
Patterson, Samuel and Gary Copeland (1994) “Parliaments in the Twenty-First
Century,” in Gary Copeland and Samuel Patterson (eds), Parliaments in the Modern
World: Changing Institutions. Ann Arbor, MI: University of Michigan Press,
pp. 1–11.
Pearse, Hilary (2005) “Geographic Representation and Electoral Reform.” Canadian
Parliamentary Review 28 (3): 26–32.
Pegnato, Joseph (1997) “Is a Citizen a Customer?” Public Productivity and Manage-
ment Review 20 (4): 397–404.
Pflanze, Otto (1990) Bismarck and the Development of Germany. Princeton, NJ:
Princeton University Press.
Pharr, Susan and Robert Putnum (eds) (2000) Disaffected Democracies: What’s Troub-
ling the Trilateral Countries? Princeton, NJ: Princeton University Press.
Pierre, Jon and Anders Widfeldt (1994) “Party Organization in Sweden: Colossuses
with Feet of Clay or Flexible Pillars of Government,” in Richard S. Katz and Peter
204 References
Mair (eds), How Parties Organize: Change and Adaptation in Party Organizations in
Western Democracies. London: Sage, pp. 332–56.
Pilet, Jean-Benoit and William P. Cross (2014) “The Selection of Party Leaders in
Comparative Perspective,” in Jean-Benoit Pilet and William P. Cross (eds), The
Selection of Political Party Leaders in Contemporary Parliamentary Democracies.
London: Routledge, pp. 447–77.
Pilet, Jean-Benoit and Bram Wauters (2014) “The Selection of Party Leaders in
Belgium,” in Jean-Benoit Pilet and William P. Cross (eds), The Selection of Political
Party Leaders in Contemporary Parliamentary Democracies. London: Routledge,
pp. 30–46.
Plasser, Fritz and Gunda Plasser (2002) Global Political Campaigning: A Worldwide
Analysis of Campaign Professionals and Their Practices. New York: Praeger.
Plott, Charles (1967) “A Notion of Equilibrium and Its Possibility under Majority
Rule.” American Economic Review 57 (4): 787–806.
Poguntke, Thomas and Paul Webb (eds) (2005) The Presidentialization of Politics.
Preston, Paul (1986) The Triumph of Democracy in Spain. London: Methuen.
Pruysers, Scott and William P. Cross (2016) “Candidate Selection in Canada: Local
Autonomy, Centralization, and Competing Democratic Norms.” American Behav-
ioral Scientist 60 (7): 781–98.
Pulzer, Peter (1975) Political Representation and Elections in Britain. London: Allen
and Unwin.
Putnam, Robert D. (2002) “Conclusion,” in Robert D. Putnam (ed.), Democracies in
Flux. Oxford: Oxford University Press.
Ranney, Austin (1962) The Doctrine of Responsible Party Government: Its Origins and
Present State. Urbana, IL: University of Illinois Press.
Ranney, Austin (1965) Pathways to Parliament: Candidate Selection in Britain. Madison
WI: University of Wisconsin Press.
Reif, Karlheinz, Roland Cayrol, and Oskar Niedermayer (1980) “National Political
Parties’ Middle Level Elites and European Integration.” European Journal of Pol-
itical Research 8 (1): 91–112.
Rentoul, John (2017) “If Emmanuel Macron wins the French election next weekend,
he would show that the centre-left is not yet dead.” Independent, April 15. http://
www.independent.co.uk/voices/emmanuel-macron-france-election-socialist-party-
a7685176.html.
Renwick, Alan and Jean-Benoit Pilet (2016) Faces on the Ballot. Oxford: Oxford
University Press.
Riordon, William (1995 [1905]) Plunkitt of Tammany Hall: A Series of Very Plain
Talks on Very Practical Politics. New York: E. P. Dutton and Co.
Roberts, Geoffrey K. (1989) “Party System Change in West Germany: Land-Federal
Linkages.” West European Politics 13: 98–113.
Roche, Daniel (1998) France in the Enlightenment, translated by Arthur Goldhammer.
Cambridge, MA: Harvard University Press.
Rokkan, Stein (1977) “Towards a Generalized Conception of Verzuiling.” Political
Studies 25 (4): 563–70.
References 205
Rosanvallon, P. (2011) Democratic Legitimacy: Impartiality, Reflexivity, Proximity.
Princeton, NJ: Princeton University Press.
Rose, Richard (1974) The Problem of Party Government. London: Macmillan.
Rose, Richard and Phillip L. Davies (1994) Inheritance in Public Policy: Change
without Choice in Britain. New Haven, CT: Yale University Press.
Rose, Richard and Harve Mossawir (1967) “Voting and Elections: A Functional
Analysis.” Political Studies XV (2): 173–201.
Rose, Richard and Derek Urwin (1969) “Persistence and Change in Western Party
Systems since 1945.” Political Studies 18 (3): 287–319.
Roth, Gunther (1963) The Social Democrats in Imperial Germany: A Study in Working
Class Isolation and National Integration. Totowa, NJ: Bedminster Press.
Ruostetsaari, Ilkka (2003) “Finland: From Political Amateurs to Political Class,” in
Jens Borchert and Jürgen Zeiss (eds), The Political Class in Advanced Democracies:
Rustow, Dankwart (1955) The Politics of Compromise: A Study of Parties and Cabinet
Government in Sweden. New York: Greenwood Press.
Sappington, David (1991) “Incentives in Principal-Agent Relationships.” Journal of
Economic Perspectives 5 (2): 45–66.
Sartori, Giovanni (1976) Parties and Party Systems: A Framework for Analysis.
Cambridge: Cambridge University Press.
Sartori, Giovanni (1986) “The Influence of Electoral Systems: Faulty Laws or Faulty
Method,” in Bernard Grofman and Arend Lijphart (eds), Electoral Laws and Their
Political Consequences. New York: Agathon Press, pp. 43–68.
Scarrow, Susan (ed.) (2002) Perspectives on Political Parties: Classic Readings. New
York: Palgrave Macmillan.
Scarrow, Susan (2013) “New Challenges of Intra-Party Democracy: Grassroots
Activists, Instant Members, and Cyber-Militants,” in Emmanuelle Avril and
Christine Zumello (eds), New Technology, Organizational Change and Governance.
Houndmills: Palgrave-Macmillan, pp. 85–101.
Scarrow, Susan (2015) Beyond Party Members: Changing Approaches to Partisan
Mobilization. Oxford: Oxford University Press.
Scarrow, Susan and Burcu Gezgor (2010) “Declining Memberships, Changing Members?
European Political Party Members in a New Era.” Party Politics 16 (6): 823–43.
Scarrow, Susan, Paul Webb, and Thomas Poguntke (eds) (2017) Organizing Political
Parties: Representation, Participation, and Power. Oxford: Oxford University Press.
Scharpf, Fritz (1999) Governing in Europe: Effective and Democratic? Oxford: Oxford
University Press.
Scharpf, Fritz and Vivien Schmidt (eds) (2000) Welfare and Work in the Open Econ-
omy. Oxford: Oxford University Press.
Schattschneider, E. E. (1942) Party Government. New York: Rinehart.
Schlesinger, Joseph (1994) Political Parties and the Winning of Office. Ann Arbor, MI:
University of Michigan Press.
Schmitt, Hermann (2014) “Partisanship in Nine Western Democracies: Causes and
Consequences,” in John Bartle and Paolo Bellucci (eds), Political Parties and Parti-
sanship: Social Identity and Individual Attitudes. London: Routledge, pp. 75–87.
206 References
Schmitter, Philippe (2001) “Parties Are Not What They Once Were,” in Larry
Diamond and Richard Gunther (eds), Political Parties and Democracy. Baltimore,
MD: Johns Hopkins University Press.
Schumpeter, Joseph (1962 [1942]) Capitalism, Socialism and Democracy. New York:
Harper and Row.
Segatti, Paolo (2014) “The Italian Election of February 2013: A Temporary Shock or a
Harbinger of a New Party System?” in Silvio Beretta, Axel Berkofsky, and Fabio
Rugge (eds), Italy and Japan: How Similar Are They? A Comparative Analysis of
Politics, Economics, and International Relations. Milan: Springer, pp. 121–36.
Sen, Amartya (2012) “The Crisis of European Democracy.” New York Times, May
23: A23.
Seyd, Patrick and Paul Whiteley (2002) Labour’s Grass Roots: The Politics of Party
Membership. Oxford: Clarendon Press.
Shaw, E. (2001) “New Labour: New Pathways to Parliament.” Parliamentary Affairs
54 (1): 35–53.
Shively, W. Phillips (1979) “The Development of Party Identification among Adults:
Exploration of a Functional Model.” American Political Science Review 73 (4):
1039–54.
Simeon, Richard (1977) Must Canada Fail? Montreal: McGill-Queen’s University
Press.
Skelcher, Chris (1998) The Appointed State: Quasi-governmental Organizations and
Democracy. Buckingham: Open University Press.
Skocpol, Theda (1992) Protecting Soldiers and Mothers: The Political Origins of Social
Policy in the United States. Cambridge, MA: Harvard University Press.
Smith, Adam (1937 [1776]) The Wealth of Nations. New York: Modern Library.
Smith, Mitchell P. (1997) “The Commission Made Me Do It: The European Com-
mission as a Strategic Asset in Domestic Politics,” in Neill Nugent (ed.), At the
Heart of the Union. Houndmills: Palgrave Macmillan.
Smithies, Arthur (1941) “Optimum Location in Spatial Competition.” Journal of
Political Economy 49: 423–39.
Sorauf, Frank (1959) “Patronage and Party.” Midwest Journal of Political Science
3 (2): 115–26.
St. Leger, M. (2009) “German court upholds tax challenge.” National Catholic Reporter,
August 28. http://ncronline.org/news/global/german-court-upholds-church-tax-
challenge, accessed 11 May 2011.
Ştefuriuc, Irina (2009) “Government Formation in Multi-Level Settings: Spanish
Regional Coalitions and the Quest for Vertical Congruence.” Party Politics 15 (1):
93–115.
Steinmo, Sven (2010) The Evolution of Modern States: Sweden, Japan, and the United
States. Cambridge: Cambridge University Press.
Stewart, David K. (forthcoming) “Primaries and the Personalization of Party Leadership,”
in William P. Cross, Scott Pruysers, and Richard S. Katz (eds), The Personalization of
Democratic Politics. London: Rowman and Littlefield.
Stewart, David K. and Keith Archer (2000) Quasi-Democracy? Parties and Leadership
Selection in Alberta. Vancouver: University of British Columbia Press.
Stone, Deborah (1988) Policy Paradox and Political Reason. New York: HarperCollins.
References 207
Stone, Jon (2015) “Tory MP who claimed moat cleaning on his expenses has been
handed a peerage.” http://www.independent.co.uk/news/uk/politics/tory-mp-
who-claimed-moat-cleaning-on-his-expenses-will-return-to-parliament-through-lords-
10474701.html, accessed December 10, 2015.
Stott, T. (1995) “ ‘Snout in the Trough’: The Politics of Quangos.” Parliamentary
Affairs 48 (2): 323–40.
Streeck, Wolfgang (1993) “The Rise and Decline of Neocorporatism,” in Lloyd
Ulman, William T. Dickens, and Barry Eichengreen (eds), Labor and an Integrated
Europe. Washington, DC: Brookings, pp. 80–101.
Strm, Kaare (1990) “A Behavioral Theory of Competitive Political Parties.” Ameri-
can Journal of Political Science 34 (2): 565–98.
Strm, Kaare, Wolfgang C. Müller, and Torbjőrn Bergman (eds) (2003) Delegation
and Accountability in Parliamentary Democracies. Oxford: Oxford University Press.
Sundberg, Jan (1994) “Finland: Nationalized Parties, Professionalized Organizations,”
in Richard S. Katz and Peter Mair (eds), How Parties Organize: Change and Adap-
tation in Party Organizations in Western Democracies. London: Sage, pp. 158–84.
Svåsand, Lars (1994) “Change and Adaptation in Norwegian Party Organizations,” in
Richard S. Katz and Peter Mair (eds), How Parties Organize: Change and Adapta-
tion in Party Organizations in Western Democracies. London: Sage, pp. 304–31.
Thayer, Nathaniel (1969) How the Conservatives Rule Japan. Princeton, NJ: Princeton
University Press.
Thorlakson, Lori (2009) “Patterns of Party Integration, Influence and Autonomy in
Seven Federations.” Party Politics 15 (2): 157–77.
Torgersen, Ulf (1970) “The Trend towards Political Consensus,” in Erik Allardt and
Stein Rokkan (eds), Mass Politics. New York: Free Press, pp. 93–104.
Törnudd, K. (1968) The Electoral System of Finland. London: Hugh Evelyn.
Transparency International (2013) Global Corruption Barometer 2013. Berlin: Trans-
parency International.
Transparency International (2016) People and Corruption: Europe and Central Asia:
Global Corruption Barometer. Berlin: Transparency International.
Tromborg, Mathias Wessel (2014) “Bringing the Median Voter Back In: The Rela-
tionship between Government Debt, Median Voter Preferences, and Welfare State
Spending.” Journal of European Social Policy 24 (2): 107–21.
Truman, David (1951) The Governmental Process: Political Interests and Public Opin-
ion. New York: Alfred A. Knopf.
United States Department of Justice (2010) “Horizontal Merger Guidelines.” Issued
August 19. https://www.justice.gov/sites/default/files/atr/legacy/2010/08/19/hmg-
2010.pdf.
Valen, Henry and Daniel Katz (1964) Political Parties in Norway: A Community
Study. Oslo: Universitetsforlaget.
Van Schuur, Hendrik (1984) Structure in Political Beliefs: A New Model for Stochastic
Unfolding with Application to European Party Activities: Proefschrift. Amsterdam:
CT Press.
Van Thiel, Sandra (2004) “Quangos in Dutch Government,” in Christopher Pollitt and
Colin Talbot (eds), Unbundled Government: A Critical Analysis of the Global Trend
to Agencies, Quangos and Contractualisation. London: Routledge, pp. 167–83.
208 References
Vincent, J. and M. Stenton (1971) McCalmont’s Parliamentary Poll Book: British
Election Results 1832–1918. Brighton: Harvester Press.
Volkens, Andrea, Pola Lehmann, Theres Matthieß, Sven Regel, Bernhard Weßels
(2012) The Manifesto Data Collection: Manifesto Project (MRG/CMP/MARPOR).
Berlin: Wissenschaftszentrum Berlin für Sozialforschung.
Webb, Paul (1992) “United Kingdom,” in Richards S. Katz and Peter Mair (eds),
Party Organizations: A Data Handbook. London: Sage, pp. 837–70.
Webb, Paul and Dan Keith (2017) “Assessing the Strength of Party Organizational
Resources: A Survey of Evidence from the Political Party Database,” in Susan
Scarrow, Paul Webb, and Thomas Poguntke (eds), Organizing Political Parties:
Representation, Participation, and Power. Oxford: Oxford University Press,
pp. 31–61.
Webb, Paul, David Farrell, and Ian Holliday (2002) Political Parties in Advanced
Industrial Democracies. Oxford: Oxford University Press.
Weber, Max (1919) “Politik als Beruf” (Politics as a Vocation). Lecture at University
of Munich. English translation: https://archive.org/details/weber_max_1864_1920_
politics_as_a_vocation.
Weir, Stuart and David Beetham (1999) Political Power and Democratic Control in
Britain: The Democratic Audit of the United Kingdom. London: Routledge.
Weeks, Liam (2008) “Candidate Selection: Democratic Centralism or Managed
Democracy?” in Michael Gallagher and Michael Marsh (eds), How Ireland Voted
2007. Basingstoke: Palgrave.
Werden, Gregory (2004) “Economic Evidence on the Existence of Collusion: Recon-
ciling Antitrust Law with Oligopoly Theory.” Antitrust Law Journal 71 (3): 719–800.
Whiteley, Paul, Harold Clarke, David Sanders, and Marianne Stewart (2001) “Turn-
out.” Parliamentary Affairs 54 (4): 775–88.
Wiesli, Reto (2003) “Switzerland: The Militia Myth and Incomplete Professionalization,”
in Jens Borchert and Jürgen Zeiss (eds), The Political Class in Advanced Democracies:
Will, George (2015) “Defining economic failure down.” Washington Post, February 5.
Wilson, Alex (2012) “Presidents and Parties in Regional Politics: Italy and
Spain Compared.” Barcelona: Institute of Political and Social Sciences, working
paper 308.
Wilson, Alex (2016) “Regional Presidents, Multi-Level Parties and Organisational
Stratarchy: The Case of Italy,” in Ferdinand Müller-Rommel and Fernando Casal
Bértoa (eds), Party Politics and Democracy in Europe. London: Routledge.
Yishai, Yael (2001) “Bringing Society Back In: Post-Cartel Parties in Israel,” Party
Politics 7 (6): 667–87.
Young, Lisa, Anthony Sayers, Harold Jansen, and Munroe Eagles (2005) “Implica-
tions of state funding for party organization.” Paper presented at the annual meeting
of the American Political Science Association, Washington, DC.
Index
Aarts, Kees 90 Bartels, Larry 59

Agius, Christine 164 Bartolini, Stefano 41, 86
Albright, Madeline 121 Barzelay, Michael 27
Aldrich, John 106 Bawn, Kathleen 111
Alford Index 83–6 Beaumont, Peter 154, 156
Allern, Elin 123, 160 Beckett, Margaret 69
Allum, Percy 156 Beer, Samuel 4, 32, 37, 169
Amery, L. S. 171 Beetham, David 120
Ames, Paul 115 Belgium 35, 40, 45, 56, 58–9, 63, 65,
Andeweg, Rudy 54 70–1, 75, 85–6, 88–9, 95, 98, 112,
Archer, Keith 178, 182 119, 132–3, 135–6, 139, 151,
Argentina 116 154–6, 160–1, 167
Attlee, Clement 94–5 Christian Democratic Party (Parti
Aubry, Martine 183 Social Chrétien; Christelijke
Audickas, Lukas 142 Volkspartij) 158
Australia 76, 85, 88–90, 95, 133, 136, Christian Democratic and Flemish
155, 167 Party (CD&V) 54
Country Party 88 Christian Social Party (Parti Social
Labor Party 88 Chrétien) 158
Liberal Party 88, 136 Greens (Agalev) 186–7
Austria 39–40, 44, 45, 63, 71, 85–6, Humanist Democratic Centre (Centre
88–9, 95, 97–9, 103–5, 112, 116, démocrate humaniste) 158
119, 132–5, 151–2, 155–7, 160, Liberal Party (Partij voor Vrijheid en
167, 176 Vooruitgang; Parti de la Liberté et
Constitutional Court du Progrès) 54, 88, 98
(Verfassungsgerichtshof ) 152 Socialist Party (Parti Socialiste;
Freedom Party (FPÖ) 135, 152, 187 Socialistische Partij) 54, 58, 88, 98
Greens (Die Grünen – Die Grüne Vlaams Bloc 151
Alternativ) 152 Volksunie 56
Peoples Party (ÖVP) 45, 88, 97, 99, Walloon Rally (RW) 56
132, 135, 139, 152, 176 Bellamy, Richard 156
Socialist Party (SPÖ) 45, 88, 97, 99, Benoit, Kenneth 88, 117
132, 135, 139, 152, 176 Bentham, Jeremy 34
Avnon, Dan 104, 105 Berelson, Bernard 162, 185
Berlusconi, Silvio vii, 67, 87
Bachrach, Peter 29 Best, Heinrich 76, 79–80
Bäck, Hanna 71 Beus, Jos de 7
Bairoch, Paul 35 Bevan, Aneurin 65, 81
Ball, Stuart 39 Bickerton, Christopher 29, 175
Barber, Benjamin 29 Biezen, Ingrid van 2, 18, 57–8, 62–4, 67,
Bardi, Luciano 55, 56 78, 104, 108, 116, 120, 140, 143
210 Index
Bille, Lars 69 Castles, Francis G. 2, 88, 101
Birmingham 32 Cato Institute 175
Bischoff, Carina 17 Central banks 16, 18, 28, 93, 145
Blair, Robert 110 Chrétien, Jean 138
Blair, Tony 67, 69, 93–5, 164 Christensen, Jrgen 121
Blechinger, Verena 140 Churchill, Winston 44, 94–5, 169
Blumler, Jay 96 Cicero 96
Blyth, Mark 6, 9, 20, 91, 164 Clancy, Paula 73
Boda, Michael 132 Clark, Joe 66
Bolleyer, Nicole 70, 143 Coalitions 4–5, 14, 43, 71, 78, 82–3,
Borchert, Jens 16, 72, 74, 76 97–100, 111–12, 132–6, 139, 143,
Bottom, Karin 72 158, 180
Bowler, Shaun 59, 118 government formulas 71, 97–9, 134–5
Brans, Marleen 75, 119 incongruent coalitions 71
Brants, Kees 96 Cognitive mobilization 15, 139
Brennan, Jason 181 Cold War 9, 30–1, 45, 139
Brougham, Henry 163 Coleman, Stephen 96
Bruter, Michael 64–5, 77 Collier, David 29
Bryan, William Jennings 186 Collignon, Stefan 156
Buchanan, James 111 Colomer, Josep 103
Buchanan, Patrick 19 Communist parties 45, 56, 83, 139
Budge, Ian 97 Comparative Manifestos Project 85, 159
Burke, Edmund 32–4, 43, 167, 170, 180 Conservative parties 41, 85–92
Butler, David 89, 140, 163 Constitutional courts 106, 117, 145
Convention on the Elimination of all
Callaghan. James 137 Forms of Discrimination Against
Campbell, Kim 138 Women 109
Canada 66, 76, 85, 88–90, 95, 110, 112, Copeland, Gary 174
115, 118, 121, 133, 138, 155, 162 Corbyn, Jeremy 69, 150, 168, 178
Canada Elections Act 20, 66, 110 Corn Laws 35
Candidate selection 70, 110 Costa Rica 116
Conservative Party 70, 113 Cotta, Maurizio 76, 79–80
House of Commons 112 Council of Europe 51, 108
Liberal Party 66, 70, 85, 88, 113, Cracknell, Richard 142
117–18, 138 Crisis of democracy 2, 23, 24, 30, 154,
Natural Law Party 110 156, 168, 181
New Democratic Party 70 Cross, William 65–6, 70, 171
Progressive Conservative Party 66, Crossman, Richard 65, 113
88, 138 Crozier, Michel 24, 44, 156, 168
Rhinoceros Party 110 Czech Republic 59, 63, 104, 133, 155, 167
Caramani, Daniele 33, 86, 175
Carson, Ben 180 Dahl, Robert 49, 129, 147
Cartel 129–32 Dalton, Russell 10, 15, 30, 83–4, 89,
Cartel party see Political party types 139, 154
Cartwright, Edward 6 D’Aspremont, C. 131
Carty, R. Kenneth 69, 143 Dathan, Matt 69
Index 211
Davies, Phillip L. 18, 163 Estonia 104, 133, 155, 167
Delaware 106 Eurobarometer 28, 167
Democracy European Union v, vi, 8–9, 28,
audience democracy 7 93, 108, 137, 149, 166, 173–4,
consociational democracies 40, 42, 98, 179, 181
112, 120, 133–4, 139, 157 European Central Bank 18, 93
liberal democracy 2, 24, 27–8, 34, 89, European Commission 16, 18
92, 104, 107, 129, 147, 173 European Court of Justice 93
pluralist model 40, 49, 147, 170, 180 European Parliament 58, 112
polyarchy 129 Maastricht Treaty vi, 8, 93, 137
socialist democracy 4 Referenda on EU treaties 137, 149, 178
Denmark 8, 56, 63, 73, 85–6, 88–90, Schengen Agreement 93
95, 98, 119, 123, 133, 135, 137, Single European Act 93
155, 160 Stability and Growth Pact 177
Christian Democrats European Commission for Democracy
(Kristendemokraterne) 159 through Law (Venice
Conservative party (Konservative Commission) 51
Folkepartei) 88, 135 European Convention for the Protection
Danish People’s Party (Dansk of Human Rights and
Folkeparti) 153 Fundamental Freedoms 109
Liberal Party (Venstre) 135 European Court of Human Rights 109,
Progress Party (Fremskridtspartiet) 56 110, 159
Social democrats Europeanization 143, 177, 178
(Socialdemokraterne) 136 Euroskepticism 173
Socialist Peoples Party (Socialistisk Experts 114, 146, 150, 162, 164–5,
Folkeparti SF) 56, 135 168–9, 174, 175, 177, 180
Depauw, Sam 185
Deschouwer, Kris 58 Fabbrini, Sergio 71
Detterbeck, Klaus 71, 143 Facebook 161
Di Rupo, Elio 54 Farneti, Paolo 40
Diamond, Peter 6 Federalist Papers 33
Diffuse support 166–7, 175 Fenno, Richard 184
Dogan, Mattei 87, 89, 90 Figueroa v. Canada 20
Döring, Herbert 177 Finer, Samuel 93, 145
Downs, Anthony 5, 33, 61, 129, 137, Finland 35, 45, 56, 63, 83, 85–6, 88–9,
163, 171–2, 178 95, 98, 104, 112, 133, 135–6, 153,
DuPont Pioneer 130 155, 160, 167, 185
Duverger, Maurice 4, 37, 103, 126 Finnish People’s Democratic League
(SKDL) 65, 135
Easton, David 166, 167 National Coalition Party (Kansallinen
Eddy, Melissa 178 Kokoomus) 88
Edinburgh 35 True Finns 153
Egeberg, M. 123 Fiorina, Carly 180
Eggers, Andrew 75 Fisher, Justin 68
Eldersveld, Samuel 69, 143 Flinders, Matthew 122
Epstein, Leon 11, 18, 41, 108 Flouzat, Denize 121
212 Index
Foxe, Ken 174 Ginsberg, Benjamin 148
France 32, 36, 41, 63, 66, 74, 76, 78, 83, Goertz, Gary 35
85–6, 88–90, 95, 102, 103, 110, Goodwin-Gill, Guy 132
121, 133, 137, 139, 151–3, 155–6, Greece 8, 62–3, 133, 135–6, 153, 155,
160–1, 167, 174, 183 167, 173, 175, 177
Bank of France 12 Golden Dawn 179
Code of Education 158 New Democracy 135–6, 179
Constitution 102, 107, 109 PASOK 135–6
Front National 110, 151–3 Syriza 153, 168, 179
La France insoumise 153 Green parties 56, 151, 187
Poujadistes 151 Grovey v. Townsend 108
Republicans 88, 152, 183
Socialists 88 Haas, Peter 166
Freedland, Jonathan 168 Hagevi, Magnus 74–5
Hainmueller, Jens 75
Gaebler, Ted 27 Halman, L. 90–1
Galbraith, John Kenneth 30 Harrison, Sarah 64–5, 77
Gallagher, Michael 65, 133, 155 Hawkins, Oliver 142
Galli, Giorgio 83 Hazan, Reuven 65, 66, 79
Galston, William A. 182 Head, Brian 113
Gauja, Anika 61, 171 Heidar, Knut 54
Germany 19, 29, 32–3, 45, 56, 62–4, 70, Heller, William 113
76, 83, 86, 88–90, 93, 98, 104, 105, Heraclitus 128
107, 133–6, 139, 143, 153, 155–6, Herfindahl-Hirschman Index 132
158, 160, 167–8, 176 Hirschman, Albert 110
Alternative for Germany (AfD) Hollande, François 183
153, 187 Holmberg, Sören 83
Basic Law 101, 107, 159, 172 Hondeghem, Annie 119
Christian Democratic Union Hood, Christopher 16
(CDU) 60, 88, 99, 135, 158, 176 Hopkin, Jonathan 67
Christian Social Union (CSU) 88, 99, 176 Hotelling, Harold 129
Conservative party 134 Houska, Joseph 39–40
Constitutional Court Hungary 63–4, 104, 126, 133, 153, 155,
(Bundesverfassungsgericht) 19, 167, 168
105, 152, 159 Jobbik 153
Green Party 176, 186 Huntington, Samuel 9
Großen Koalitionen 99
National Liberals 134 Iceland 65, 85, 89, 90, 95, 98, 104, 133,
Social Democratic Party (SPD) 48, 56, 155
88, 99, 135, 176 Ideal types 7, 18, 26, 31, 124, 126, 128–9,
Godesberger Programm 48 130, 136, 139–41, 144
Weimar Republic 104, 168 Ideology 9, 14, 17, 21, 23, 27, 37, 38, 43,
Zentrum 158 47–8, 50, 64, 68, 74, 81–2, 89,
Gezgor, Burcu 64 91–2, 94, 96, 102, 108, 110–11,
Gilbert, W. S. 173 124, 129–30, 164, 166, 180, 183
Gingrich, Newt vi, 14 Ignazi, Piero 154
Index 213
India 114 Jay, Antony 28, 148
Inglehart, Ronald 15, 88, 139 Jennings, Ivor 33
Institute for Fiscal Studies 157 Jensen, Hanne Nex 73–4, 119, 123
International Covenant on Civil and Jouvenel, Robert de 78, 100
Political Rights 109 Jun, Uwe 76
International IDEA 51
International Monetary Fund 9, 18, 169, Kaid, Lynda Lee 96
174, 179 Kaiser, André 145
Internet v, 60, 67, 114, 143, 161, 186, 187 Kalyvas, Stathis 39, 159
Invernizzi Accetti, Carlo 29, 175 Kamarck, Elaine C. 182
Ireland 56, 62–3, 66, 70, 73, 75–6, 85–6, Karvonen, Lauri 104
88–90, 95, 98, 99, 105, 120, 133, Keith, Dan 62
137, 176 Kennedy, John F. 173
Fianna Fáil 88, 99, 135, 175 Key, V. O., Jr. vii, 12
Fine Gael 56, 88, 99, 120, 135, 176 Keynesianism 91
Greens 175 Kirchheimer, Otto 10, 14, 45–6, 47,
Labour Party 99, 135, 176 51–2, 103, 129, 170
Progressive Democrats 99, 135, 175 Kitschelt, Herbert 17, 20–1
Iron law of oligarchy 17, 37, 165 Klaus, Josef 97
Irwin, Galen 54 Knudsen, Tim 123
Israel 104, 114, 140, 167, 179–80 Kohl, Helmut 164
Kadima 180 Kolodny, Robin vi
Labor 180 Koole, Ruud 20, 42, 54, 57, 64, 107
Likud 180 Kopecký, Petr 67, 73, 116, 119, 120, 123,
Italy 40, 45, 56–7, 63, 66, 70, 75–6, 143
85–9, 95, 98–9, 103, 105, 112, Kosiara-Pedersen, Karina 67, 143
133, 140, 143, 151, 153, 155–6, Kreisky, Bruno 97
160, 167, 175 Kriesi, Hanspeter 23
Associazioni Cristiane dei Lavoratori Krouwel André 10
Italiani 39
Chamber of Deputies 105, 112 Laakso, Markku 132
Communist Party (PCI) 56, 85, 88, Laclau, Ernesto 170
99, 187 LaPalombara, Joseph 49
Democratic Party 66, 88 Latvia 132, 133, 155
Five Star Movement (MoVimento 5 Lautsi and Others v Italy 159
Stelle) 153 Laver, Michael 87–8
Forza Italia vii, 67, 87–8 Law, C. M. 35
Club Azzurro 67 Lawson, Kay 125
Rifondazione Comunista 56 Lazar, Marc 71
Socialist Party (PSI) 56, 99 Lazarsfeld, Paul 162
Tangentopoli scandal 157 Le Pen, Marine 152–3, 174
Uomo Qualunque 151 Leach, Jim 121
Lebaron, Frédéric 121
Jackson, Andrew 103, 111 Legitimacy
Janda, Kenneth 114 diffuse support 166–7, 175
Japan 76, 113 output legitimation 149, 162–8
214 Index
Leterme, Yves 54 Members of parliament 6, 32, 33, 36, 38,
Levinson, Marc 167 58–9, 66, 69, 73–8, 80, 101, 106,
Levitsky, Steven 29 112–15, 140, 142, 157, 172
Lewis-Beck, Michael 166 allowances (as a form of
Liberal parties 41, 85–9, 92, 98 compensation) 77, 112
Liegel, Barbara 119 salaries 74–8, 112–13, 121
Lijphart Arend 81, 112, 133, 139, Members of the European Parliament
145–7, 166 66, 75, 153
Lindberg, Leon 149 Merkel, Angela 178
Linz, Juan 74 Mershon, Carol 113
Lipinski, Daniel 184 Michel, Charles 54
Lipset, Seymour Martin 5, 39, 82–4, Michels, Robert 17, 37, 53, 76, 141, 165, 171
125, 158 Mill, John Stuart 34, 129, 179
Lithunia 104, 133, 155 Miller, Gary 6
Longley v. Canada 110 Milton, John 129
Lorwin, Val 40–1 Minkin, L. 79
Luxembourg 40, 56–7, 85, 89, 95, 98, Mitchell, G. J. 8
133, 135–6, 155, 167 Monsanto 130
Communists (Kommunistesch Partei Monti, Mario 175
Lëtzebuerg, KPL) 56 Moro, Aldo 99
Greens 136 Mossawir, Harve 39
Lynn, Jonathan 28, 148 Mosse, B. 27
Mudde, Cas 156
Mac Con Uladh, Damian 154, 156 Mueller, Dennis 172
Madison, James 33 Müller, Wolfgang 2, 10, 43, 57, 104,
Majone, Giandomenico 16, 146 112, 119
Major, John 94–5 Murphy, Grainne 73
Manin, Bernard 7, 29 Mylonas, Harris 179
Marcy, Wiiliam 111
Market concentration 132, 154 Namier, Lewis 32
Marketplace of ideas 129 Narud, Hanne Marthe 74
Athenian Agora 129 Nash equilibrium 13, 82, 137
free market 130, 132, 137 Nassmacher, Karl-Heinz 55, 140
Martin, Paul 112 Netherlands 32, 35–6, 39–40, 42, 45,
Martin, Shane 185 54, 56, 62–4, 67, 71, 75, 86,
Marx, Karl 25, 48, 49 88–90, 95, 98, 112, 120, 133–6,
Masket, Seth 184 137, 139, 150, 153, 155–8, 160,
Matthews, Felicity 122 161, 167, 176
May, John D. 65, 68, 183 Antirevolutionary Party (ARP) 84, 158
McClosky, Herbert 185 bishop’s letter 81
McConnell, Mitch 184 Catholic Peoples Party (KVP) 84, 158
McKenzie, Robert 37, 39, 44, 48 Christen-Democratisch Appèl 135,
McKibbin, Ross 41 150, 152, 158, 176
McRae, Kenneth 70 Christian Historical Union (CHU) 84,
McTernan, John 69 158
Mélenchon, Jean-Luc 153 ChristenUnie 136, 159
Index 215
Dutch Reformed Church 90 Pajala, Antti 83
Greens (Groen Links) 150 Panebianco, Angelo 37, 57, 59, 107
Labor Party (PvdA) 20, 88, 135, Paolucci, Caterina 67
150, 176 Pappalardo, Adriano 42
Liberal Party (Volkspartij voor Vrijheid Paris 35, 78
en Democratie, VVD) 88, 98, Parliamentarization 57, 73
150, 158 Parsons, Talcott 82
Lijste Pim Fortuyn 136 Party finance 11–12, 18–19, 68,
Netherlands Institute for Multiparty 114–17, 118
Democracy 51 short money 115
Partij voor de Vrijheid vii, 67–8, 88, Party government 2–4, 7, 22–4, 29–31,
150, 152–3, 175 33, 52, 73, 81, 101, 119, 126, 132,
party law 68 144–5, 147, 154, 162, 166, 168–70,
school law of 1920 82 172, 175, 181, 185
Staatkundig Gereformeerde Partij Party identification 8, 10, 39–40, 50, 154,
84, 159 166, 167
Neumann, Sigmund 38 Party laws 10, 11, 68, 103–6, 107, 109,
New Public Management 16, 27 115, 117–18, 172
New Zealand 76, 85, 88–9, 95, 112, 133, Vermont 108
155, 167 Party manifestos 84–7, 91–4, 137
Labour Party 88 Party membership 1, 3, 8, 10, 12, 15, 17,
National Party 88 20, 31, 36–45, 48, 50–3, 61–70,
Noble, T. 90 107–8, 114, 139, 141, 143, 147,
Non-Departmental Public Bodies 150, 154, 160–1, 165, 171, 178,
119, 122 181–2, 185
Nordlinger, Eric 47 Party organization
Norris, Pippa 96 activists 17, 21, 41, 48, 50–1, 60, 65,
Norway 41, 56–7, 63, 74, 76, 85–6, 88–9, 69, 70, 74, 79, 122, 143, 150, 160,
95, 98, 99, 123, 133, 135, 153, 155, 164, 182, 183, 184
160, 167 ancillary organizations 16, 37–9, 42,
Conservative Party (Høyre) 88 49, 139
Labor Party (DNA) 88 candidate and leadership
Progress Party selection 65–6, 68–70, 80, 108,
(Fremskrittspartiet) 151, 153 109–11, 121, 143, 180, 182–3, 185
NOS-journaal 96 one-member-one-vote ballots 66,
Nossiter, Adam 157 68, 70
primary elections 65–6, 81, 108, 182–5
Obama, Barack 121, 167, 174 internal party democracy 51, 108, 139,
O’Donnell, James 84 171–2, 173, 185, 187
Olson, Mancur 118 party in central office vii, 12, 17–18,
O’Malley, Eoin 120 25–6, 36–8, 43–4, 49, 53–5, 57–61,
OPEC 131–2, 134 70, 72, 139, 148, 165, 171
Orth, Samuel 105 party in public office vii, 12, 15, 17, 23,
Osborne, David 27 33, 36, 37–8, 43–4, 47–9, 53–61,
Oscarsson, Henrik 83 64–6, 68–72, 78–9, 139, 143, 164,
Ostrogorski, Moise 1 171, 178
216 Index
Party organization (cont.) Political Party Database 58
party on the ground vii, 12, 15, 17–18, Political party types
24, 25, 32, 36–8, 43–4, 48–9, 53–5, anti-party-system-parties 1, 19, 24,
57–62, 64–5, 67–70, 72, 139, 143, 100, 151, 156, 178, 181, 185–7
171, 178 business firm party 67
party staffs 12, 37, 55–6, 58–9, 61, 73, cartel party 7, 13–28, 124–50
77, 79, 112, 121 catch-all party 8, 9, 14, 16, 19, 21,
stratarchy 69–72, 141, 143 25–6, 43–52, 53–4, 79, 103, 124,
three faces of party organization vii, 126–9, 141, 159, 170
25, 38, 53–80 caucus party see elite party
Party systems electoral professional party 107
cartel of parties vii, 13–14, 21, 103, elite party 31–4, 38, 48, 53, 125, 128,
133, 134–8, 139 141, 147–8
cordons sanitaires 83 franchise model 69, 72
effective number of parties 86, 132, mass party vii, 4, 8, 9, 14, 19, 23, 25,
133, 155 35–43, 44, 46–52, 53, 61, 67, 72,
fragmentation 110, 154 80, 107, 114, 124, 126–9, 139, 141,
threshold for representation 19, 110 147, 158, 164, 171, 180, 185, 187
Partyness of government and classe gardée 45, 50, 83, 126
society 23–4, 38 encapsulation 37, 39, 44, 141, 187
Patronage 6, 73, 103, 105, 110–12, mainstream parties 1–2, 6–7, 12, 16,
118–23 19, 22, 23, 27, 47, 56, 57, 61, 69,
lottizzazione 103, 112, 157 74, 77, 87, 88, 93, 94, 100, 123,
spoils system 111, 157 127, 132, 137, 139, 143–4, 149–50,
Proporz 103, 112, 157 152, 156–60, 166, 173–81, 186–7
Patterson, Samuel 174 modern cadre party 107
Pegnato, Joseph 27 parties as public utilities 18, 78,
Perot, Ross 19 108, 163
Personalization 60, 71, 82, 94–6, 183, 185 Populism 1, 16, 19, 22–4, 27, 29, 77, 83,
Pflanze, Otto 134 151–7, 162, 168, 173–7, 178,
Pharr, Susan 30 180–1, 184–7
Pierre, Jon 67 Portugal 35, 59, 62–3, 66, 104, 114, 133,
Pilet, Jean-Benoit 54, 65, 66, 185 135, 155, 167
Plasser, Fritz and Gunda 59, 79 Social Democrats 99
Plato 128 Poutou, Philippe 156
Plott, Charles 137 Prandi, Alfanso 83
Poguntke,Thomas 61, 79 Presidentialization 61, 82, 94
Poland 59, 104, 110, 133, 154, 155, 167 Preston, Paul 8
Policy Convergence 21, 82, 83–94, 150, Principal-agent models 2–7, 21, 96, 126,
160, 164, 171–2 132, 147, 162
Brexit 137, 149 agents 4–7, 20, 79–80, 109, 111, 144,
Butskellism 45 146–8, 163, 172, 174, 182–3
median voter 5, 82, 87, 137, 172, 184 Prodi, Romano 56
Political careers 6, 12, 16, 59, 65, 72–8, Professionalization 17, 26, 60, 76, 79,
110, 121, 163 114, 140, 163
Political class 11, 72–4, 77, 100, 123 Public goods 15, 27, 118, 131
Index 217
Pruysers, Scott 70 Scarrow, Susan 30, 33, 36, 41, 58–9, 62,
Pulzer, Peter 84 64, 67, 68, 73, 161
Putnam, Robert 30, 90 Scharpf, Fritz 45, 149
Schattschneider, E. E. 2, 28, 30, 181
Quangos 119–20 Scheingold, Stuart 149
Scherlis, G. 119
Rahat, Gideon 65, 66, 79 Schlesinger, Joseph 5, 61, 106
Ramsden, John 107 Schmidt, Vivien 45
Ranney, Austin 69, 140, 172 Schmitter, Philippe 8
Rational choice 6, 87, 107, 128, 140 Schroeder, Gerhard 93, 164
Nash equilibrium 13, 82, 137 Schumpeter, Joseph 1, 5, 33, 61,
prisoner’s dilemma 21 105, 106
Reagan, Ronald 164, 166 Segatti, Paolo 175
Régimes censitaires 34, 36, 44 Sen, Amartya 154, 156
Regulatory state 16, 144–8, 180 Seyd, Patrick 64
Reif, Karlheinz 79 Shaw, E. 70
Referenda 93, 137, 149, 178 Shively, W. Phillips 39, 90
Religion 22, 39–40, 46, 51, 81–2, 84, Sieberer, Ulrich 104
90, 158–60 Simeon, Richard 156
Rentoul, John 178 Sinowatz, Fred 97
Renwick, Alan 185 Skelcher, Chris 120
Renzsch, Wolgang 71, 143 Skocpol, Theda 12, 78
Residential mobility 160–1 Slovakia 62–3, 132–3, 155
Riis, O. 91 Slovenia 133, 155, 167
Riordon, William 105 Smith v. Allwright 108
Roberts, Geoffrey 71 Smith, Adam 130
Roche, Daniel 35 Smith, Alfred E. 181
Rogers, Will 185 Smith, Mitchell P. 18
Rokkan, Stein 5, 39, 42, 82–3, 90, Smithies, Arthur 129
125, 158 Social class 4, 31, 81, 83
Rosanvallon, P. 170 communist subculture 40
Rose, Richard 2, 18, 39, 81, middle class 47, 76, 89–90, 141
84, 163 social mobility 90
Roth, Gunther 38 working class 35–6, 47, 89–90, 125
Royden, Agnes Maude 84 Social cleavages 9, 15, 23, 39–42, 45–6,
Ruostetsaari, Ilkka 74 49–57, 80, 82–4, 86–7, 91, 96, 98,
Rustow, Dankwert 40–1 125–6, 139, 143, 158–9
Rutte, Mark 150, 156 frozen cleavages 42, 82–3
ontzuiling 46
Saatchi & Saatchi 60, 96 race 84
Sanders, Bernie 168, 183 segmentation 8, 39–41, 47, 49, 90, 125
Sapin, Michel 121 verzuiling 42
Sapiro, Virginia 90 Social democratic parties 38–9, 45, 56,
Sappington, David 6 81, 85–9, 91–5, 98, 139, 160,
Sartori, Giovanni 33, 110 164, 186
Saward, Michael 2 Sorauf, Frank 119
218 Index
Spain 59, 62–3, 66, 70–1, 76, 90, National Action 151–2
104–5, 133, 137, 153, 155, 160, Social democrats 98–9
167, 175 Swiss People’s Party 99, 152
Ley Orgánica 106
Podemos 153, 168 Technocracy 9, 28, 29, 113, 146, 162,
St. Leger, M. 90 164–6, 170, 174–5, 177, 180–1
Staatkundig Gereformeerde Partij against Thatcher, Margaret 107, 164, 167
the Netherlands 109 Thayer, Nathaniel 113
Staiger, Uta 156 Thorlakson, Lori 70
Ştefuriuc, Irina 71 Todosijevic, Bjorn 90
Steinmo, Sven 94 Torgersen, Ulf 41
Stenton, M. 32 Törnudd, K. 104
Stepan, Alfred 74 Trade unions 36–7, 39–40, 41, 51,
Stewart, David 178, 182 66, 114
Stokes, Donald 89, 163 Transparency International 167
Stone, Deborah 113 Trilateral Commission 168
Stone, Jon 157 Trinidad & Tobago 114
Stott, T. 174 Troika 18, 175
Streeck, Wolfgang 149 Tromborg, Mattias Wessel 137
Strm, Kaare 2, 43, 173 Truman, David 49, 147
Strömbäck, Jesper 96 Trump, Donald 150, 152, 180, 181
Stronach, Belinda 113 Tsipras, Alexis 179
Suffrage vii, 32, 34–6, 39, 44–5, 47, 81, Tullock, Gordon 111
101, 125 Twitter 161
Sundberg, Jan 72
Svåsand, Lars 49 Union density 160
Sweden 8, 35, 40, 56–7, 62–3, 67, 71, United Kingdom 4, 8, 24, 29, 32–5, 41,
74–6, 84–6, 88–9, 90, 93–5, 98, 44–5, 56–7, 60, 62–6, 68, 71, 73–7,
116, 133, 135–6, 151, 153, 155, 83, 84–6, 88–90, 93, 95–6, 98, 102,
160, 161, 164, 167 104, 112, 115, 119, 122, 133, 135,
Conservative party (Moderata 137, 140, 142, 153–7, 161, 167,
samlingspartiet) 56, 88, 93–4, 170, 176, 178
136, 164 British Audit Commission 121
Greens 136 Church of England 84
Liberal Party (Liberalerna) 56, 136 Commissioner for Public
Social democrats (Sveriges Appointments 122
socialdemokratiska Conservative Party 39, 41, 44–5, 48,
arbetareparti) 88, 136, 164 60, 66–9, 75, 88, 94, 96, 107, 122,
Sweden Democrats 137, 140, 142, 149, 157, 169–71,
(Sverigedemokraterna) 151, 153 176, 178
Switzerland 32, 40, 56, 63, 73, 76, 85–6, Conservative Working-Men’s
89, 93–5, 98, 132, 133–5, 139, Clubs 39
151–2, 155, 160, 167 National Union of Conservative and
Democrats 152 Unionist Associations 44
Federal Council 98, 132, 152 Primrose League 44, 67
Liberal Party 98 Corn Laws 35
Index 219
House of Commons 104, 115, 140, Venice Commission see European
142, 149, 178 Commission for Democracy
Labour Party 38, 40, 41, 48, 54, 60, through Law
64–7, 69–70, 84, 88, 94, 96, 105, Verge, Tània 160
115, 120, 122, 137, 140, 142, 150, Vermont 108
157, 161, 176, 178 Vincent, J. 32
Clause IV 47 von Beyme, Klaus 11, 77
Liberal Democrats 68, 85, 88, 149 von Treitschke, Heinrich 33
Liberal Party 34, 56, 149 Voter turnout 1, 8, 10, 24, 154–6, 176,
National Health Service 122 179–80, 182
Reform Act of 1867 32, 44
Scottish National Party (SNP) 105, 149 Wallace, George 175
United Kingdom Independence Party Wattenberg, Martin 10, 30, 154
(UKIP) 153 Wauters, Bram 54
Whig Party 34 Webb, Paul 41, 56, 59, 61–2, 79
United States v–vii, 8, 12, 19, 30, 45, 51, Weber, Max 26, 65, 76
60–1, 69–70, 76, 83–4, 102–3, 104, Weeks, Liam 70
105–6, 108, 111–12, 117, 121, Weimar Republic 104–68
130–1, 152, 154, 167–8, 174–5, Weir, Stuart 120
180–2, 185 Welfare state 8, 15, 45–6, 91–2, 94,
Democratic Party vi, 14, 19, 84–6, 88, 163–4
106, 108, 121, 175, 181, 183–4, 186 Whiteley, Paul 24, 64, 154
Department of Justice 132, 154 Widfeldt, Anders 67
Federal Communications Wiesli, Reto 73
Commission 121 Wilde, Oscar 34
Federal Election Campaign Act 106 Wildenmann, Rudolf viii, 2, 101
Federal Trade Commission 121 Wilders, Geert vii, 67, 150, 156, 174
Republican Party vi, 14, 19, 88, Will, George 167
106, 121, 150, 152, 168, 175, Wilson, Alex 70–1, 143
180, 183–4 Wilson, Harold 137
Urwin, Derek 84 World Bank 9, 174
World Trade Organization 8–9, 28, 93,
Valen, Henry 41, 90 174, 181
van der Kaap, Harry 90 World Values Survey 89–90
van Holsteyn, J. J. M. 64
van Kersbergen, Kees 159 Yishai, Yael 128, 179–80
van Praag, Philip 42, 96 Young, Lisa 18
Van Schuur, Hendrick 79 Yumak and Sadak v Turkey 110
Van Thiel, Sandra 120
Vartiainen, Hannu 6 Zeiss, Jürgen 72, 74, 76

(Comparative Politics (Oxford University Press) ) Katz, Richard S. - Mair, Peter - Democracy and The Cartelization of Political Parties (2018, Oxford University Press)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Comparative Politics (Oxford University Press) ) Katz, Richard S. - Mair, Peter - Democracy and The Cartelization of Political Parties (2018, Oxford University Press)

Uploaded by

Copyright:

Available Formats

OUP CORRECTED PROOF – FINAL, 28/5/2018, SPi

DEMOCRACY AND THE CARTELIZATION

OTHER TITLES IN THIS SERIES

Multi-Level Electoral Politics

RICHARD S. KATZ AND PETER MAIR{

This book represents the culmination of a project that began at an ECPR

1.1 Simple principal-agent model of democracy 3

3.1 Party leadership in Belgium, the Czech Republic, Poland,

xiv List of Tables

2 Democracy and the Cartelization of Political Parties

THE SIMPLE PRINCIPAL-AGENT MODEL

In its simplest form, the principal-agent model of democracy in a parliamentary

Party or parties in Ministry State

F I G U R E 1 . 1 Simple principal-agent model of democracy

4 Democracy and the Cartelization of Political Parties

Segment 1 Ministry State

F I G U R E 1 . 2 Mass party principal-agent model with three social segments/parties

F I G U R E 1 . 3 Downsian principal-agent model with three parties and coalition government

6 Democracy and the Cartelization of Political Parties

8 Democracy and the Cartelization of Political Parties

THE CARTEL THESIS

We initially arrived at the idea that new patterns of relationships were

10 Democracy and the Cartelization of Political Parties

12 Democracy and the Cartelization of Political Parties

CARTEL PARTIES AND A PARTY CARTEL

14 Democracy and the Cartelization of Political Parties

MAKING SENSE OF CARTELIZATION

16 Democracy and the Cartelization of Political Parties

18 Democracy and the Cartelization of Political Parties

20 Democracy and the Cartelization of Political Parties

to prevent the emergence of challenges from outside the cartel . . . Thus in

22 Democracy and the Cartelization of Political Parties

24 Democracy and the Cartelization of Political Parties

These developments then raise three important questions, or research agenda,

26 Democracy and the Cartelization of Political Parties

28 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties

30 Democracy and the Cartelization of Political Parties

Since party is central to our current understanding and learned practice of

The Rise and Decline of Parties 31

THE ELITE PARTY

32 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 33

34 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 35

36 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 37

38 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 39

40 Democracy and the Cartelization of Political Parties

A party in Sweden as elsewhere in Europe is not only a political appar-

The Rise and Decline of Parties 41

Divisions and patterns of segmentation in other European polities were not

42 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 43

THE CATCH-ALL PARTY

44 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 45

46 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 47

48 Democracy and the Cartelization of Political Parties

The Rise and Decline of Parties 49

50 Democracy and the Cartelization of Political Parties